Vous êtes sur la page 1sur 645


Toward a Theory of Hllman Food Habits

Toward a Theory of Human Food Habits

[iijiiJ ~


Temple University Press, Philadelphia 19122 Copyright © 1987 by Temple University. All rights reserved Published 1987 Printed in the United States of America The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences-Permanence of Paper for Printed Library Materials, ANSI 239.48-1984

Library of Congress Cataloging-in-Publication Data
Food and evolution. Includes bibliographies and indexes. 1. Food habits. 2. Human evolution. 3. Nutrition-Social aspects. 4. Man, Primitive-Food. I. Harris, Marvin, 1927II. Ross, Eric B. [DNLM: 1. Food Habits. GT 2860 F686] GN407. F65 1986 306 86-5773 ISBN 0-87722-435-8 (alk. paper)

Although now substantially revised, the initial drafts of the papers in this volume were presented at the 94th Symposium of the Wenner-Gren Foundation for Anthropological Research at Cedar Key, Florida, October 23-30, 1983. On behalf of all the participants, the editors wish to thank the foundation and its staff for their support and advice. We were especially aware of our debts to Lita Osrnundsen, the foundation's director of research. The editors are also deeply grateful to the staff of Temple University Press, especially to Jane Cullen, Jennifer French, and Jane Barry for their heroic production and copyediting feats.

Introduction Part I. Theoretical Overview 1. An Overview of Trends in Dietary Vlriation from Hunter-Gatherer to Modern Capitalist Societies ERIC B. ROSS 2. Foodways: Historical Overview and Theoretical Prolegomenon MARVIN HARRIS Part II. Bioevolutionary Antecedents and Constraints 3. Primate Diets and Gut Morphology: Implications for Hominid Evolution KATHi\RINE MILTON 4. Omnivorous Primate Diets and Human Overconsumption of Meat WILLIAM]. HAMILTON III 5. Fava Bean Consumption: A Case for the CoEvolution of Genes and Culture SOL()MON H. KATZ Part III. Nutritional and Biopsychological Constraints 6. Problems and Pitfalls in the Assessment of Human P. L. PELLETT Nutritional Status 7. Psychobiological Perspectives on Food Preferences and Avoidances PAUL ROZIN 8. The Preference for Animal Protein and f'at: A Cross-Cultural Survey H. LEOr~ ABRAMS, JR.












9. Biocultural Consequences of Animals Versus Plants as Sources of Fats, Proteins, and Other Nutrients LESLIE SUE LIEBERMAN Part IV. Pre-State Foodways: Past and Present


10. The Significance of Long-Term Changes in Human
Diet and Food Economy MARK N. COHEN

261 285

11. Life in the "Garden of Eden": Causes and
Consequences of the Adoption of Marine Diets by Human Societies DAVID R. YESNER

12. The Analysis of Hunter-Gatherer Diets: Stalking an
Optimal Foraging Model BRUCE WINTERHALDER


13. How Much Food Do Foragers Need?

341 357

14. Aboriginal Subsistence in a Tropical Rain Forest
Environment: Food Procurement, Cannibalism, and Population Regulation in Northeastern Australia DAVID R. HARRIS

15. Ecological and Structural Influences on the
Proportions of Wild Foods in the Diets of Two Machiguenga Communities ALLEN JOHNSON and MICHAEL BAKSH


16. Limiting Factors in Amazonian Ecology
KENNETH R. GOOD Part V. The Political Economy and the Political Ecology of Contemporary Foodways


17. Loaves and Fishes in Bangladesh

427 445

18. Animal Protein Consumption and the Sacred Cow
Complex in India K. N. NAIR



19. The Effects of Colonialism and Neocolonialism on the Gastronomic Patterns of the Third 'Vorld RICHARD W. FRANKE 20. Stability and Change in Highland Andean Dietary Patterns BENjAlvlIN S. ORLOVE 21. Social Class and Diet in Contemporary Mexico GRI~TEL H. PELTO 22. From Costa Rican Pasture to North American Hamburger Mi\RC EDELMAN Part VI. Discussion and Conclusions 23. The Evolution of Human Subsistence






24. Biocultural Aspects of Food Choice GEORGE ARMELAGOS Mterword About the Contributors Glossary Name Index Subject Index

595 601 607 613 625


THIS BOOK RESULTS FROM AN INTER_DISCIPLINARY EFFORT TO ADvance our understanding of why human beings in differing times and places eat what they do. It begins, at the most fundamental level, with the collective view of the editors and other contributors that knowledge and comprehension of human foodways, and the web of practices and beliefs associated with them, must depend upon our seeking general principles and recurrent processes beneath the immediate appearance of a worldwide confusion of seemingly capricious preferences, avoidances, and aversions. Once this decision is made, however, a complex set of explanatory strategies and options still remains to be explored and integrated, since the knowledge we have of human food customs and practices derives from data collection that has traditionally been dispersed among varied specialties and theoretical strategies. We cannot claim that all the relevant disciplines or all the salient levels of analysis and perspective are represented in the chapters that follow, nor do we presume that this work encompasses an adequate representation of those that are. But we hope at least to have helped to broaden the general scope of inquiry beyond the horizons of any single viewpoint, while still maintaining what we emphatically regard as a commitment to a nomothetic approach. The disciplinary perspectives of the contlibutors to this volume range over primatology (Hamilton, Milton), nutrition (Pellett, Lieberman), biological anthropology (Armelagos, Katz), archaeology (Yesner, Cohen, Roosevelt, D. Harris), psychology (Rozin), and agricultural economics (Nair). Although cultural anthropologists predominate numerically, they too offer a great diversity of insight and information based on their varying professional interests and, in particular, their wide spectrum of regional specializations: Bangladesh (Lindenbaum), Amazonia (Johnson and Baksh, Good, Ross), Paraguay (Hawkes), Canadian sub-arctic (Winterhalder), Southeast Asia and Africa (Franke), Mexico (Pelto), Costa Rica (Edelman), Peru (Orlove), and Europe (Ross). In attempting to integrate the diversity of disciplinary viewpoints that these scholars represent, the editors chose an evolutionary framework as the only suitably broad yet coherent and unifying one available to us. In its biological dimensions, at least, it seemed self-evident that the core of human dietary practice, all subsequent embellishment aside, must be regarded in terms of the



emergence of the hominidae and the co-evolution of human diet and our physical potential for cultural behavior. It seems likely, for example, that hunting for vertebrates, increased meat consumption, and expanded tool use were implicated in the evolutionary processes that led to the expansion and reorganization of the australopithecine brain and to the development of Homo's unique capacities for consciousness and semantic universality. There is, at least, little doubt that, throughout most of the Pleistocene, the evolution of biological repertoires and the evolution of behavioral repertoires were closely intertwinedand that diet is one domain where the intersection was particularly noteworthy. With the appearance of Homo sapiens, if not earlier, however, a progressively greater independence-or lag-between biological and cultural selection reduced the rate and incidence of gene-culture co-evolution. Radically different modes of production, accompanied by massive changes in food habits, emerged in the later phases of human prehistory and throughout subsequent history without any discernible evidence of related changes in gene frequencies. Increasingly, behavior associated with the procurement, distribution, and consumption of food came, like the rest of human behavior, to be propagated through learning rather than genetic replication. And although selection based on consequences for reproductive success continued to operate, it was increasingly supplemented, if not displaced, by selection based on the more immediate consequences for the satisfaction of biopsychological needs and drives. Though the feedback between these two levels of selection became increasingly indirect and delayed, the biological level still cannot be excluded from our attempt to understand general as well as particular aspects of the evolution of foodways. Indeed, in a small number of cases such as that of fava bean (see Chapter 5) and milk consumption, specific preferences and avoidances continue to be associated with genetic polymorphisms found with varying frequencies among different populations. In evolutionary perspective, however, most of the great changes in human diet can be more readily associated with shifts in modes of production that are not in tum linked to such genetic variations. The transition from upper paleolithic to neolithic modes of production, for example, generally involved a shift from narrow reliance on animal foods to broad-spectrum regimens in which the consumption of domesticated tubers and grains gained ascendancy over meat and other animal foods (pastoral modes of production, of course, followed a divergent trajectory). The next great general evolutionary changes in foodways may be associated with the rise of archaic agro-managerial states whose dense, socially stratified populations were dependent on one or two staple grains and which maintained distinctive consumption patterns for elites and commoners. The further evolution of imperial state systems with massive potentials for trade and great capacities for modifying their habitats through public works doubtless increased such class or caste distinctions in dietary practice and gave



rise to new rural/urban and core/periphery distinctions. These effects of political-economic evolution and of ever more formidable political-ecological integration finally attained global proportions with the emergence of the capitalist world system, leading in our own times to a return to highly carnivorous diets for privileged regions and classes at the cost of impoverished diets and often de facto vegetarianism in dependent and underdeveloped areas. The contributions to this volume have been arranged with these broad evolutionary considerations in mind. Following the editors' theoretical overviews, Parts II and III deal with the biological, nutritional, and psychological factors that reflect species-specific and/or population-specific consequences of human genetic repertoires. Part IV deals with food patterns associated with pre-state sociocultural systems as revealed through both archaeological and ethnographic researches. And Part V concerns itself with foodways in contemporary state-level societies, with emphasis upon the consequences of underdevelopment and participation in the capitalist world system. Needless to say, it is impossible for any single volume to provide a thoroughly comprehensive treatment of so conlplex a subject as the evolution of human food habits or to reach any definitive theoretical outcome. What we have hoped to do, however, is to provide a guide and a framework for much needed future investigation, and an incentive for others to join in that necessarily collaborative enterprise. M. H. and E. B. R.


Theoretical Overvie"T
THE TWO ESSAYS THAT FOLLOW SHARE AN EXPLICIT MATErialist strategy and are addressed specijically to the question of the general determinants offood preferences and avoidances. They range over a variety ofpre-state and state-level foodways, highlighting food t'ractices that have generally been regarded as beyond the pale of nomothetic approaches or whose cost-benefit signijicance is in dispute. The epistemological basis for distinguishing idealist from materialist approaches to foodways rests on the separation of data obtained through emic operations from those obtained through etic operations. Emic foodways data result from eliciting operations in which the participants' sense of what people eat or ought to eat, and the symbolic signijicance of food preferences and avoidances, dominate data collection. On the other hand, etic foodways data do not necessarily require eliciting operations and are reported in a data language whose units and categories are imposed by the observers (e.g., calories, proteins, costs and benefits). Beyond the separation of emics from etics, materialist approaches to foodways start with the assumption that puzzling dietary habits are the outcome ofdeterminative processes in which biopsych0 logical, technological, economic, demographic, and environmental factors predominate. These infrastructural processes account for the evolution of distinctive forms ofstructures and superstructures. Once such structures (e.g., domestic and political organization) and superstructures (e.g., religious and symbolic systems, Philosophies, aesthetic standards) are in place, they ofcourse exert an influence over all aspects ofsociallzfe, includingfoodways. Religious food taboos, for example, have a distinctive role to play in the maintenance offood habits. But recognition that structural and superstructural features react back upon infrastructure does not lessen the distinction between materialist and idealist approaches or justify taking reful?e in an eclecticism that is incapable of weighing one causal component against another or of stating the conditions under which now infrastructure, now superstructure, achieves dominance. The restraints imposed by infrastructure upon structure and superstructure remain dominant in the determinative processes that lead to continuity or change in food-


1. Theoretical Overview

ways: foodways that acquire adverse etic cost-benefit balances will tend to be selected against; foodways that have favorable etic cost-benefit balances will tend to be selected for. Participants' emic valuations of foodways arise from infrastructure. Major changes in infrastructure cause major changes in foodways and their emic valuations. Changes in emic valuations change major foodways, but only when such changes are favored by infrastructural conditions. As both essays stress, the balance of etic costs and benefits that provides the cultural and biopsychological selection pressures for and against particular foodways often differs markedly according to age- and sex-related status roles and social strata. Hierarchies based on sex, class, ethnicity, and other distinctions are usually associated with favorable cost/benefits for some status roles but unfavorable cost!benefits for others. Where such conditions prevail, the study of foodways must form part of the study ofpolitical economy and political ecology. As in contemporary state societies and their neocolonial dependencies, what people eat is often what they are allowed to eat or obliged to eat as a consequence of their subordination to the material priorities of ruling classes and corporate elites.



An Overview of Tren<is in Dietary Variation from Hunter-Gatherer to Modern Capitalist Soc:ieties

THE STUDY OF VARIATION AND C:HANGE IN HUMAN DIETSespecially in that most enigmatic dimension of dietary custom, food preferences and avoidances-is important in several respects. In a general way, it compels us to confront and challenge the possibilities and limits of cultural explanation (see Ross 1980), while, more specifically, it enables us to begin to seek (and hopefully to formulate) generalizable and predictive principles in a domain of culture where we are routinely led to believe that such principles are unlikely to apply. For it is precisely in the matter of dietary customs that the concept of culture has been most consistently invoked to suggest that, at the heart of what seems the most material and practical of human affairs, there lies an ineluctable core of arbitrary, fortuitous, or irrational thought. The seemingly inexhaustible variety and range of human dietary patterns thus has been taken to represent quintessential evidence of the inexorable power of the human mind, through the device of culture, to transcend the constraints of material and historical circumstances. The implications of such arguments are not merely academic, for they suggest that where such circumstances coincide with impoverished diets, the power to alter and improve them must be ulltimately as arbitrary and fortuitous as the forces that originally induced them. vVhat basis does such a view provide for people to devise a program to change and improve their lives? If, as some have suggested (see Sahlins 1976: 171), contemporary cultural patterns are, in the end, the arbitrary or random variants of mysterious structures in the human mind, then change too must be arbitrary, and the avenues by which certain definite desired ends might be achieved are beyond effective reach.


I. Theoretical Overview

Our attempt here to explore the forces and circumstances that appear to shape the diverse patterns of human diet is, in contrast, inherently an effort to explain how human behavior in general diversifies in intelligible, if not predictable, ways. But, ultimately, it must take into account a more practical end: we do not study human dietary customs merely because they are interesting, but because at the end of the day they help to define the quality of life of real people. Variation in what people eat reflects substantive variation in status and power and characterizes societies that are internally stratified into rich and poor, sick and healthy, developed and underdeveloped, overfed and undernourished. An anthropology that, by one methodological device or another, reduces such features of any social system to arbitrary reflections of the human mind only entraps itself in a self-indulgent relativism that precludes all likelihood of raising a coherent critical voice. It is a necessary requisite of the perspective I advocate that diet be viewed within a historically formulated understanding of any given social system, as an evolutionary product of environmental conditions and of the basic forces, especially the social institutions and social relations, that effectively determine their use. In this paper, a preliminary effort has been undertaken to suggest some of the implications of such an approach.

Hunters, Horticulturalists, and Dietary Preferences
Much recent attention in the field of human diet has been given to the lifestyles of pre-industrial populations. Ross (1978a) marked one of the first attempts to redirect such work away from an almost exclusively symbolic approach to one that envisaged diet as part of a larger adaptational process. At much the same time, others were beginning to apply concepts of animal ecology-in particular, optimal foraging theory (see Winterhalder and Smith 1981)-to extant populations of human hunters. To date, however, the blessings of such applications have been decidedly mixed. One reason, at least, has been that the populations studied have not been economic isolates. As a specific kind of cost-benefit model, foraging theory is most successful where energy is the principal behavioral constraint. But no living human societies are immune to the additional constraints imposed by other human beings, which range from the relatively benign to the invasive. And few if any hunting populations are ever as isolated as those who study them may like to believe. An additional problem is perhaps more distinctively theoretical. One of the enduring issues involved in dietary custom-and one of decided relevance to the general question of dietary variation-is that of taboos (see Harris 1977,


1. Overview of Trends in Dietary Variation

1979; Ross 1978a). Yet optimal foraging theory has yielded little insight into such matters-indeed, may be considered to have confounded them-and the reasons why suggest some formidable limitations that I will briefly suggest here. In neotropical regions such as Amazonia, several factors have been concurrently implicated in the avoidance of certain game animals (Ross 1978a). First, there is the biomorphological character of the animals themselves: comparative evidence suggests that larger game anirrlals-such as the tapir-are extremely vulnerable to predation, much as their even larger paleolithic antecedents were to prehistoric hunters. Second, there is the question of alternative resources: where aquatic resources exist, if they are productive, they tend to be more efficient and stable sources of food than game animals and to claim a greater share of labor time. Third, much depends on the degree of sedentism of the human population, since this has a profound effect on the distribution and density of game. The interaction of these factors can be expressed by the following principle: access to productive aquatic resources inclines to an increase in sedentarism, since such resources are relatively immobile; increased sedentarism will usually mean greater development of horticulture (Carneiro 1968), while a more sedentary, horticultural lifestyle will differentially diminish access to the spectrum of available game animals. Thus, small ones, which reproduce rapidly or are commensal in nature, tend to be found near settlements, regardless of continued hunting, while larger animals such as tapir and deer require treks into ever deeper forest. If aquatic resources are lucrative enough, such animals will rarely be hunted and may eventually be regarded as inedible; indeed, if aquatic resources are unusually plentiful-for example, estuarine or lacustrine-even some smaller animals may not be targeted as a source of food. Thus, Wilbert has observed of the Warao of the Orinoco delta "that some of the nutritionally most valuable game animals, especially tapir, and here and there also deer and paca, are traditionally tabooed" (1972:68-69), while Basso has noted that the Kalapalo, vv-ho inhabit the rich river and lake network of the Upper Xingu, "regard virtually all land animals ... as disgusting and refuse to eat them" (1973: 14, 16; (~arneiro 1970; Murphy and Quain 1955:29). In contrast, where the aquatic resources that so enhance the opportunities for sedentary life are not present, hunting tends to remain necessarily the paramount subsistence activity, insofar as the intake of proteins and fats is concerned. An example is provided by the i\che of eastern Paraguay, who still spend a considerable portion of their year as hunters. At such times there is no horticultural activity, and fishing is relatively insignificant "because the streams that flow through [their] territory are poorly supplied with fish" (Clastres 1972: 155). Thus, in contrast to the riverine Kalapalo, the Ache seem to exclude virtually no potential animal food from their diet (Hill and Hawkes 1983).


I. Theoretical Overview

In between these two extremes lies a continuum of degrees of exclusion, determined by various mixes of hunting and fishing productivity, which in turn depend on such variables as settlement size, type of horticulture, exposure to trade, and warfare. It is reasonable to suppose that hunters such as the Ache will regard the tapir as edible, if our delineation of the determinants of preferences and avoidances is more or less an accurate one. Indeed, as Hawkes and Hill have recently commented, "Several tapir were shot during the study period, and we recorded 10 man hours spent in pursuit of them without success. If one tapir were to have been killed during this period it would have been the number one ranked item in the optimal diet" (ibid.). But, in fact, in their account of the "optimal" Ache diet, tapir does not appear. As they explain, "The resource rankings of this model say nothing about the quantitative importance of a resource to optimal foragers. High ranked items may be so rarely encountered that they contribute only a very small proportion of the diet" (ibid.). 1 In the case of tapir, they seem to have made so little contribution to the Ache diet-indeed, none at all during the research period-that they were not even included in the model's rankings of what Ache preferences ought to have been. Initially this may seem vaguely to express the remarkably ambiguous role that the tapir occupies as the largest, yet one of the most elusive, game animals. But the optimal foraging model, at least in this instance, contains a notable irony, for Hill and Hawkes tell us that the tapir would have been the highestranked animal had it made any contribution to the diet at all. This suggests, as we have already seen, that the model is much less interested in the dynamics of hunting tapir in general than in the productivity of one successfully hunted tapir, however rare an occurrence that might be. The question is whether native preferences would arise out of the former or the latter. The anthropologists' "optimal" diet model at least suggests that the model's preference is based upon the latter. But this creates a significant dilemma, for although the model would then view the tapir as the optimal prey, ethnographic observation indicates that, as among so many other neotropical groups, it is the white-lipped peccary that is the primary target of Ache hunting (Hill and Hawkes 1983). This contradiction compels us to affirm that our primary aim is to comprehend the forces that shape ethnographically recorded preferences and real dietary consequences-not to displace them with a model that contains its own peculiar and often compelling logic. Thus, optimal foraging tells us little if anything about why the tapir is frequently excluded from the diet among many hunter-horticulturalists. The reason for this is simple: as we have previously suggested, the tapir's sensitivity to human predation renders it progressively more scarce the more sedentary a community of hunters becomes. But by the admission of its own advocates, optimal foraging theory disregards the question of an animal's abundance or scarcity (Winterhalder 1981:93-94). On these terms, the tapir, once admitted


1. Overview of Trends in Dietary Variation

to the model, would never fall out of the diet-a conclusion clearly contradicted by evidence. It can only be said that this may not lllatter too much when considering exclusively hunting peoples, who, being mobile, can minimize the scarcity of such animals by following them without being compelled to confine themselves to a convenient distance from a stable home-base. But, certainly, once there is any measure of sedentarism to consider, the contradictions of the model begin to diminish its credibility. Indeed, paradoxically, what began as a putatively materialist model ends by emulating the lnore abstract, symbolic models it presumably sought to displace in its lack of due consideration for the dynamic environment in which cultural behavior operates. That cautionary lesson is not to be restricted to pre-industrial populations.

Calories, Hunting, and Dietary Breadth
One of the great differences between the diets of hunting peoples and those of hunting-horticultural populations hinges on their respective sources of calories. Among the former, the harvest of game, fish, wild plants, and insects not only embodies human energy investment but also constitutes the sole source of calories in the diet-excepting whatever flows in through non-traditional channels. For that reason, foraging societies lTLay be more readily analyzed with energy-based models. On the other hand, where horticulture is also a constituent of the economy, it not only induces a greater degree of sedentarism, but also provides more secure sources of calories. As this occurred, for example, during the Neolithic, the role of energy as a limiting factor in human activity gave way to other factors such as fats and proteins (Gross 1975; Reidhead 1980). The difference may readily be seen in a comparison of the Ache and the Achuara Jivaro of Peru: among the former, hunting supplies an estimated 80 percent of dietary calories (Hawkes, Hill, and O'Connell 1982:385), whereas among the horticultural Achuara 74 percent are derived from manioc and plantains (Ross 1976; 1978a:4). This in tum has profound implications for dietary strategies, for it means that where Ache hunting patterns must reconcile energy inputs and outputs fairly efficiently, Achuara hunting is effectively subsidized by horticulture and can be undertaken at an energetic loss if that proves necessary to bring in adequate supplies of other essential nutrients. On the other hand, because among strictly foraging populations hunting is also a source of calories, such groups frequently manifest an apparently excessive intake of meat. This has been a source of some confusion. Hawkes, Hill, and O'Connell (1982:385), for example, have observed that the rather large meat harvest of the Ache is "quite surprising in view of recent generalizations


I. Theoretical Overview

about lowland South America and low-latitude hunters in general." In fact, such surprise derives from an essential misunderstanding: the generalizations to which they refer were concerned largely with hunter-horticultural populations (Gross 1975; Ross 1978a). That noted, the scale of Ache hunting returns corresponds to what we know of other foraging peoples, among whom calories are in general the principal limiting nutrient and for whom meat is often the major source of energy (see Speth and Spielmann 1983 for a discussion of some of the implications of this for nutrition and subsistence strategies). For the same reason, foragers often are disposed to favor animal species that are particularly rich in fat-whether it is beaver among the Amerindians of the Canadian boreal forest (Berkes and Farkas 1978: 161; Winterhalder 1981) or tapir and coleoptera larvae among the Ache (Clastres 1972). As a result, it has perhaps not been uncommon for foraging populations, in a diverse assortment of biomes, to have a "surprisingly" high protein intake-in order to amass sufficient calories. Thus, to take them once more as an example, the Ache average 150 grams of protein per capita per day during their hunting treks (Hawkes, Hill, and O'Connell 1982:385). On the other hand, high game harvest may not necessarily guarantee caloric sufficiency, and there is little doubt that boreal forest hunters were under considerable energy stress (Berkes and Farkas 1978: 162; Winterhalder 1981:67-68)-a situation that undoubtedly facilitated their growing dependence upon purchased food during the post-contact period (Feit 1982). But we may be at least entitled to speculate that it is almost intrinsic to the energy-limited economy of foragers that they should tend to over-predate. If so, this may help to explain the severe game depletions of the Pleistocene, which paved the way for more diversified and sedentary lifestyles during the Neolithic. The increasingly sophisticated interpretation of the archaeological record suggests that the transition to the Neolithic was accompanied by a fairly general decline in dietary quality, evidenced in stature and decreased longevity (see Chapter 10; Buikstra 1983). There seems certainly to have been a decline in the availability of quality protein, previously provided by hunting, and an increase in the consumption of starchy plant foods. It may well have been during this period that sedentary groups first evidenced the meat craving that is reported among so many hunting-horticultural peoples today. It has been suggested that such a craving-a preference for meat as the exemplary foodembodies a form of cultural motivation, since hunting is often a problematical and unrewarding task. But there may also be a physiological component. Judith and Richard Wurtman (1983) have described, for example, the quite different effects on brain chemistry of diets high in protein or carbohydrates. Whereas a diet with a high protein-to-carbohydrate ratio leads to a lowering of brain serotonin levels and elevates the desire for carbohydrates, one lower in protein and higher in carbohydrates-the general pattern for most horticultural popula-


1. Overview of Trends in Dietary Variation

tions since the Neolithic-increases serotonin levels and induces a "craving" for protein. It is interesting to consider that, to some degree at least, the excess consumption of meat that characterizes many contemporary industrial populations may have its roots in the revolutionary transformation of the human consumption patterns that the rise of horticulture initiated. Most horticultural populations, subsisting on diets relatively high in carbohydrates, may have an elevated desire for protein-rich foods, but there are, as we have noted, practical limits to the satisfaction of such ends as long as protein is largely derived from wild animals. With the development and eventual expansion of animal domestication, however, new opportunities opened up, and in comparatively recent times, in advanced industrial nations, this has meant that although consumption of calories has remained high, the desire for increased protein intake has been able to be met. This, on the other hand, as we will see in a later section, has been achieved only through far-reaching social and economic developments that have involved an unprecedented level of integration of diverse local cultural systems, and has meant an intensification of economic, social, and political stratification, locally, nationally, and on a global scale.

Hunting, Taboos, and the Market Economy
We have observed that among foraging populations, the level of intake of protein may, in fact, be an artifact of the quest for calories. By the same token, the contribution of various animals to the diet may be in part a function of their utility beyond consumption as food: for exalnple, where they provide skins for clothing, as with deer among the pre-contact Archaic Indians of Michigan (Keene 1981: 186). This, once more, suggests that it would be erroneous to regard the motives underlying all elements in the diet as entirely circumscribed by dietary considerations. This is, perhaps, even more the case among post-contact hunters and hunter-horticulturalists, whose dietary preferences (and even avoidances) are frequently as much affected by outside influences and pressures as by cultural traditions. These exogenous factors are often disregarded when diet is considered, since an overwhelmingly "cultural" perspective on food customs tends to impel researchers to look inward, toward indigenous ideological templates. But, as Leacock has aptly noted, "in most instances, [even] peoples with a gatherer-hunter heritage have not lived solely as gatherer-hunters for a long time" (1982: 160). (For example, although the Yanoama Indians of the Venezuelan rain forest have been widely regarded as a relatively unacculturated population, their staple crop is the plantain-a I~uropean introduction.) Thus, the


1. Theoretical Overview

Ache have been presented in various essays as if they personified an authentic paleolithic lifestyle (e. g., Hawkes, Hill, and O'Connell 1982). But, in fact, it is practically impossible (if, obviously, theoretically practicable) to comprehend the dynamics of Ache foraging without taking into account that the group under study spends around half the year in a Catholic mission, where, in between hunting treks, its members are "under the supervision of the mission staff. . . . they grow manioc, sugarcane, com, and sweet potatoes, and they keep a few pigs, goats, chickens, and burros," as well as being provided with "additional resources in the form of milk, sugar, rice, flour, noodles, and salt" (Hawkes, Hill, and O'Connell 1982:381-82). It is difficult to believe that this has not had a serious effect on foraging strategies. If nothing else, it must have reduced the frequency of hunting over the course of a year, with the probable result that pressure on game has slackened (but see Chapter 13). Like the Ache, most pre-industrial peoples have, over the last few centuries, been incorporated in varying measure into an ever expanding market economy, a process that has influenced their customary behaviors-including diet-in ways that we are only slowly beginning to appreciate (Leacock 1981:39-62; Tanner 1979). One serious way has been through pressure to exploit resources for the production of non-food commodities. Thus, in Canada, though beaver was already exploited before the arrival of Europeans, increasing involvement by Indians such as the Cree in the fur trade seems to have increased the role of the beaver as a food source (Feit 1982:380; Tanner 1979:60-61; Winterhalder 1981: 87). This, however, had at times the additional effect of nearly decimating the species (Feit 1982:390; Tanner 1979:61), thus affecting diet in the opposite direction. In Amazonia today, there are few Amerindians who do not participate to some degree in such commercial activities as selling or bartering skins and timber. In Peru and parts of Ecuador, for example, jaguar and peccary skins in large numbers find their way to the entrepot city of Iquitos, from which point they proceed to the United States and Europe (Ross 1983). At present, for a group such as the Achuara, peccaries constitute over half of the total fresh weight of all game harvested; and although there is no doubt of their traditional significance as a source of food (Ross 1979), their central importance in recent decades as a source of valuable credit with local river traders undoubtedly induces hunters to kill more of them than nutritional requirements alone might warrant. To that extent, if protein intake among the Achuara is higher than is strictly necessary in dietary terms (Ross 1978a), it would be wrong to draw the inference that this is evidence of the abundance of tropical forest game (a conclusion recently drawn by Chagnon and Hames 1979). It may only reflect commercial pressures. Such outside influence has worked in a diversity of ways, as the Achuara case suggests. The Achuara today tend to regard such animals as tapir, deer,


1. Overview of Trends in Dietary Variation

and capybara as inedible (Ross 1978a). But, as with other pre-industrial populations, it is impossible to regard Amazon hUlnan ecology apart from the historical processes of the last centuries as the Amazon was drawn into the net of a global capitalist economy (Ross 1978b). There is, for example, some reason to believe that in the late nineteenth century Achuara communities were considerably larger than they are today and that they hunted and ate deer and tapir (Izaguirre 1925: 175-78; Steward and Metraux 1948: 619, 623). An increase in commercial contacts by the tum of the century seems to have altered subsistence patterns, however. First, guns were acquired. Although they may have been adopted as an instrument of warfare-which was endemic among the Achuara and other Jivaroan groups-they tended in general to displace traditional weapons. The acquisition of such shotguns in tum encouraged a more dispersed pattern of smaller settlements (a trend reinforced by new epidemic diseases) and a decline in communal hunting, a change that tended to discourage the pursuit of larger animals. Furthennore, since the use of guns depended upon access to ammunition through the rather unreliable agency of river traders, the ascendancy of such hunting instrumlents tended to be accompanied by an increasing measure of discrimination in targets, which, at least in the Achuara case, seems to have intensified the trend toward reliance on small animals and fish (Ross 1978a). The Achuara are certainly not unique in this regard. Although there may be some debate about whether traditional or new technology is more effective for neotropical hunters (see Hames 1979; Ross 1978a; 1979; Yost and Kelley 1983), it seems fairly clear that the effort required to afford and/or secure ammunition for their guns has introduced a level of risk into Amerindian hunting that in many instances has led to a significant contraction in the scope of hunting activities and, hence, in the range of contributions that hunting makes to the diet. Thus, Goldman has written of the riverine Cubeo of the Colombian Amazon:
Hunting may have been more prominent in the past, when only native weapons, spears and blowguns using curare-tipped darts were fashionable. Today the Indians hunt with muzzle-loading shot-guns that only a few possess because of the expense and trouble. A hunter is not disposed to take risks with his small store of shot, powder and percussion caps. (1963:57; see also Jackson 1983:46)

In some cases the introduction of new technology reinforced a pre-existent seasonal scheduling. Thus, in many riverine areas of the Amazon it is commonly the case that a marked seasonality in water levels dramatically affects game and fish densities. In the high-water periods, for example, aquatic resources decline severely; fish, in particular, become very scarce, and among some communities-as Stocks (1983: 246) reports for the Cocamilla of the


1. Theoretical Overview

Huallaga floodplain in eastern Peru- "there is a relative and consciously perceived strain on subsistence." Among the Shipibo on the Ucayali River, flood waters cover as much as 90 percent of their catchment zone. This may restrict fishing, but at the same time it tends to concentrate the animal population in a fairly confined land surface along the natural river levees (Behrens 1981: 19495). It is during this season, then, that "the Shipibo increasingly invest in shotgun shells and try their luck hunting. Less affluent Indians buy powder and shot to operate their ancient muzzle-loaders" (Hoffmann 1964:265). At this time of year, moreover, hunters "exploit otherwise rarely eaten game such as armadillo, coatimundi, majas [paca], and tapir" (Behrens 1981: 195)-a reminder that edibility, and preferences and avoidances, are essentially processual. If guns reinforced a pre-existent hunting strategy as in this instance, they were also able to curtail certain traditional targets. Sometimes this was because of an initial increase in the kill-rate or an intensification of the rate of game dispersal, which reduced the long-term yield of some species in a given region. But, in other cases, some animals were simply less efficiently harvested with guns. Thus, at the tum of the century, Up de Graff observed that in Jivaro country
the big monkeys are the hardest game to bring down with shot-gun or rifle. I have shot them to pieces with a Winchester, until their entrails actually fell to the ground, before they have fallen. Even when dead, very frequently they remain suspended by the tail, and the tree has to be climbed before they can be recovered. But the poisoned dart from a blow-gun ... kills them within two minutes. (1923:213)

Curare, being a muscle-relaxant, loosened the dying animal's hold on the branches, allowing it to drop to the ground to be readily claimed by a hunter. In some cases the introduction of guns seems, in this way, to have effected a transformation in the concept of the comestibility of such animals. Thus, as Hill and Hawkes (1983: 185) write of capuchin monkeys: "The Ache consider these monkeys as game (the Ache have hunted only with bows until very recently), whereas neighboring Guarani-speaking Indians, Paraguayans and Brazilian peasants, who all hunt with shotguns, do not consider these monkeys as prey (or even as edible in some cases)." A particularly interesting example of how ecological conditions have affected concepts of edibility, and how resource use and dietary custom have both been altered under outside, commercial influence, is provided by the Miskito Indians of Nicaragua's eastern coast (the ethnographic present is the pre-Sandinista period). They have access to "a wide range of biotopes," which, being unrestricted by hostile neighbors-in contrast to the situation in so many parts of Amazonia-they have been able to exploit quite successfully. But living immediately by coastal waters that include "the largest sea turtle feeding grounds in


1. Overview of Trends in Dietary Variation

the Western hemisphere" (Nietschmann 1972b:71), their use of terrestrial game resources has tended to be very selective-much as we would be led to expect from our earlier discussion (see Chapter 11; Chisholm, Nelson, and Schwarcz 1983)-in terms of both species hunted and hunting frequency. The Miskito, in fact, derive about 82 percent of their total annual harvest of meat from aquatic resources, with most hunting occurring during the months of highest rainfall when turtling declines (compare the riverine Shipibo, who derive about 76 percent of annual animal flesh from fish and similarly intensify hunting when fishing becomes seasonally restricted). That hunting is relatively unimportant through most of the year and that "more than 65 percent of the active adult ITlen concentrate their meat-getting activities only on turtling" (Nietschmann 1972a: 59) are two factors that undoubtedly have contributed to the continuing productivity of hunting zones. As most game animals have not been subjected to persistent predation, the Miskito have been enabled to hunt in a highly selective fashion, which has meant that the more elusive and problematical targets have tended to be disregarded. Predictably enough, the tapir was traditionally regarded as inedible (Loveland 1976:81). Increasingly involved in commercial exchange, which has meant catching turtles for the sale of their meat and shell (employed in the manufacture of combs, jewelry, etc.), and in need of cash to maintain their economy, which has become dependent on non-traditional techniques of turtling and fishing, the Miskito hunters have, however, been compelled to change their strategy. Commercial motives have become intertwined with purely dietary considerations. As Nietschmann has observed: "Miskito hunters and fishermen are focusing on animals with a high market potential in the village. . . . populations of green turtles, white-lipped peccary, and white-tailed deer are receiving additional pressure from human populations because of their taste preference and marketable potential" (1972a:60, 63). As a result, in "response to the decline of other meat sources" (Nietschmann 1973: 1] 2), particularly green turtles, the tapir has reentered the Miskito diet.

Protein Restriction and the Differential Costs of Dietary Pressure
The transition to horticulture led, in many regions, to new problems of resource availability, not least with regard to high-quality protein. This point has been widely noted in the recent anthropological literature (e.g., Speth and Scott 1984). In considerations of the varied human responses to this particular problem, most attention has been fixed on subsistence strategies; by com-


I. Theoretical Overview

parison, little regard has been paid to social responses that often had as much effect on dietary outcome in actual consumption by real individuals. Clearly this is not a matter that is easily assessed from the archaeological record, except in terms of certain material sequelae that express states of nutrition-related health. In this regard there is evidence that the development of horticulture, although it may have led to an increase in birthrate, actually was responsible for a rise in child mortality-perhaps due to lower consumption of quality protein and a rise in diseases associated with greater sedentarism (Hassan 1975:44). There is also reason to suspect that women suffered a reversal in quality of life. More children would, for example, have entailed greater labor costs for women (Ember 1983:296), which-along with more frequent pregnancies-would have meant increased physical stress, particularly as observations from recent and contemporary horticultural societies suggest that men rarely share more than a fraction of the energetic costs of child rearing, though they may disproportionately benefit from the economic advantages of more offspring. Such studies also indicate that women bear a greater amount of the burden of diminished consumption. Indeed, this suggests an important possibility in regard to the rise of horticulture itself. A major force behind increased efforts to domesticate plants and animals was the depletion of wild game during the late Paleolithic (Butzer 1971; Hayden 1984; Martin 1985). As this occurred, the likelihood is that the equality of distribution that typically characterizes foraging populations (Leacock 1981; Lee 1982) gave way before an attempt by men to monopolize animal flesh. (This has been observed in recent times by Lindenbaum among the Fore of highland New Guinea, where it induced women to adopt the practice of supplementing their own diet with the flesh of their deceased relatives [1979:24].) If women were thus disproportionately stressed as wild resources were depleted, there is reason to suppose that they were also more likely to take the initiative in a variety of ways in intensifying production: that is, in engaging in and developing precisely those subsistence activities that, in aggregate, we describe as the rise of horticulture. Nonetheless, in most of the pre-industrial horticultural societies that we know about today, there is dietary stress, and it is rare that this is suffered equitably through some unitary, group accommodation; summaries of average per capita intake disguise the fact that work effort, intake, or both are apportioned among the population according to such criteria as age and gender, with adult men being the typically favored group. (This generalization can probably be extended to most industrialized societies as well.) Undoubtedly the withholding of food in varying degrees from certain categories of individuals, such as neonates, must be regarded as a kind of attenuated infanticide-one that seems to have a long-standing role in human and, indeed, in hominid and pre-hominid history. The tendency to restrict the diet of adult females is more ambiguous in intent and more complex in its ramifications,


1. Overview of Trends in Dietary Variation

especially when it applies to those who are pregnant or lactating. Yet it is relatively common to find taboos or some kind of dietary restrictions-frequently regarding the consumption of protein-rich foods-that are invoked precisely during these periods. If implemented \vith any consistency, such customs would in many instances jeopardize the mother's health or her capacity to bear healthy offspring. Thus, Sharma (1955) has associated the "greatly restricted diet" of many Burmese mothers with a high rate of infantile beriberi. On the other hand, such prohibitions often exist where resources are problematical, in which case the outcome of a new pregnancy or the survival of a new infant must be weighed against the well-being of older children and of the mother herself (see Chapter 2). To consider the latter for the moment: there is widespread evidence of an association between reduced diet in pregnancy and smaller infants; although this may put the neonate at risk, it also tends to reduce the risk of childbirth (Shorter 1982: 165)-no inconsequential consideration in pre-industrial societies. Gender variation in diets has probably had a long history, if we may judge by research among non-human primates and human hunter-gatherers. Harding, for example, in 1,032 hours of observations of free-ranging olive baboons at Gilgil, Kenya, recorded 47 incidents of these baboons killing and eating small animals. Although such activities played a very small role in their total subsistence, consumption was largely restricted to adult males and generally seemed to reflect male statuses within the troop (1973, 1974). Among chimpanzees studied by Teleki and others at Gombe National Park, although actual killing was virtually monopolized by adult males (Teleki 1975: 149-50, 154), there was elaborate and extensive sharing, which, although it seems to have been associated with adult male status far less than among Gilgil baboons, was certainly structured along gender lines. As Teleki observes, "As pursuit is undertaken almost exclusively by adult males, who in tum become the owners of large portions, females (and to a lesser extent subadults) become highly dependent upon mature males for this particular item" (1975: 153). Since the overall predation rate is relatively low and thus the nutritional significance of such food is small, there is reason to view such meat consumption patterns as an instrument of social relations (Wrangham 1977). Among humans too, food distribution is ernbedded in the prevailing pattern of social relationships, both reflecting and helping to reinforce status and power. Even if hunting-gathering societies have tended, in the main, to be fairly egalitarian, the likelihood is that a sexual division of labor has necessarily entailed certain dietary divergences along gender lines. How early on such a division of labor evolved-or even precisely why it did-is too complicated an issue to discuss here, and it is, at any rate, a largely speculative question. (For example, one recent explanation of why fen1ales do not hunt is derived from contemporary evidence of amenorrhea among Western women runners; it sug-


1. Theoretical Overview

gests that because "the human manner of hunting involves outrunning prey with endurance," it would be adaptive for females not to hunt, as participation in such activities would have impaired their fertility [Graham 1975:811]. However likely this argument may be, it remains hypothetical, though it is probably the case that the sexual division of labor, which largely defines hunting as a male activity, is based on some such biosocial factor involving fertility.) For the present, we can only note that, with the evolution of a pattern of labor in which males monopolized the hunt, the likelihood arose of dietary differences between men and women-if only because the former were likely to consume a certain percentage of their catch while away from home, just as women were apt to do while gathering or, later on, while engaged in horticultural tasks. Thus, Jellife and co-workers observe of the Hadza, a foraging group in northern Tanganyika, that "the diet of the two sexes differs greatly as much of the food is eaten as soon as it is obtained-the men eating on the spot small animals or carrion" Oelliffe et al. 1962:909). A similar situation is reported among the pastoralist Karimojong of eastern Africa:
Among the Karimojong during the dry season, young men tending the camp herds (at some distance from the home settlement) may subsist wholly on animal products, while the children and older adults at the settlement will rely on only plant foods and some milk. During the wet season, more animal foods will be available to those who remain at the settlement, and camp herdsmen will have greater access to cultivated plant foods. Hence at any given time males are likely to be consuming more proteinrich animal foods than are females. (Little and Morren 1976: 59-60)

Among many hunting-horticultural societies, perhaps because pressure on game resources is greater than among foragers, the distribution of meat within communities tends to reflect status to a considerable degree: that of men over women and that among men. Thus, among the Achuara, when families apportion out cuts of meat among their relations, it is likely that nothing will be given to the family of a man who is absent from the village. Among the Yanoama of the southern Venezuelan rain-forest, Good (Chapter 16) reports that meat in most cases is distributed to men, who in tum allocate portions to the women and children. Larger portions tend to go to the more important men, while lower-status men may receive none, especially in the larger villages where game is more sparse and demand is greater. Because male status is associated with hunting ability, moreover, the families of low-ranking men suffer doubly as far as meat consumption is concerned. This, in turn, may have decided demographic consequences. First, it imposes a special risk to women, who evidence a relatively high mortality during their reproductive years (Neel 1977: 163), probably because nutritional stress renders them more vulnerable than men to disease (Chagnon 1974: 160). Second, it affects the rate of infanticide, since,


1. Overview of Trends in Dietary Variation

when food is scarce, women will be more likely to regard a neonate as competing with themselves and their older children. (Thus, headmen, with greater access to resources, tend to have offspring vvith a sex ratio indicating a significantly lower rate of female infanticide than is found for other males [Chagnon 1979].) Whether gender-oriented food restrictions or prohibitions in other societies have similar consequences is difficult to ascertain; few studies have made any systematic effort to assess either the degree to which such beliefs are actually translated into practice or their effects. It is nonetheless notable that they are frequently meant to be applied to women of child-bearing age and they almost universally encompass nutritionally important protein sources (Bolton 1972; Katona-Apte 1975; Ogbeide 1974; Trant 1954; Wellin 1955). In some cases, at least, this may have the effect of diverting such foods, which are often scarce, to adult males in the event that they become available (Trant 1954)-a diversion that is commonly justified by the argument that the male workload is greater, even though the woman may be pregnant and working. The effect if such dietary limitations are, in fact, enforced is that "women may be in a constant state of protein drain," as Ogbeide has argued for MidWest state in Nigeria (1974:215), or, perhaps, suffering from anemia (Maher 1981:81). The same effect would follow, hovvever, where prohibitions are not necessarily explicit, but where social relations within the family are conditioned by the family's ties to the wider society, in which male status is regarded as more privileged. In India, "as one moves up in the scale of income or expenditure classes, the consumption of animal protein also increases significantly" (Chapter 18). There, because women tend to work for less than men or are unwaged and thus control little income, they have less status within the household and are presumed to have less claim to consume such food (or to need it less). Thus, within the low-income rural Indian household, one finds a continuum of dietary privilege, with the woman generally situated at the bottom. According to Katona-Apte:
At meals, the wife feeds her husband first, then the children, boys before girls; only then does she eat. Often she eats separately with her daughters. If available, the major portions of such nourishing foods as meat, fish, egg, or milk, and sweets, are served to the males, because it is believed that they need it for strength or growth, and often not much is left over for the females. It is not unusual to find households where the women are vegetarians, but the fnales are not. Vegetarianism among females may be rationalized on religious grounds, thus leaving more (or all) of the high protein foods for the males. (1975:45)

A similar situation is reported from Berber-speaking communities in Morocco, where


I. Theoretical Overview

men eat before women and children and consider prized food items such as meat to be their prerogative. Women and children learn to refuse meat, and on fonnal occasions, to eat what the men leave. . . . If they are guests [women] will often swear that they have eaten already, and if they are not, that they are not hungry. One woman used to assure her fellow diners that she preferred bones to meat. Men, on the other hand, are supposed to be exempt from facing scarcity, which is shared out among women and children. (Maher 1981:80-81)

In this case the results have been noted: high mortality among children and malnutrition among adult females (ibid.: 81). As I have noted, the origins of such dietary divergence may lie deep in human prehistory, but such degrees of subordination of women's diet to the prerogatives of male consumption seem to have emerged largely with the evolution of ranked societies, as a woman's status came to depend on that of her husband or other male kin. For example, "Bemba women dispense food as a family service that redounds to the husband's stature. . . . Among the Mae Enga, women's labor furnishes produce that is consumed by the pigs which are distributed in political negotiations by men" (Leacock 1981: 159). This trend reaches its height in class societies, particularly in the case of modem industrial capitalism, with the "structural separation of the public economy of capitalism and the private economy of the home" (Davis 1981:229)the contrast between paid and unpaid work-with women disproportionately relegated to the latter, to producing children, facilitating the reproduction of the labor force (in the case of the working class),2 and generally supporting the position of the men on whom their own livelihoods depend. 3 As Rowbotham has commented, if the woman "wants to 'improve' herself, she has in fact to 'improve' the situation of her husband. She has to translate her ambition into his person" (1983:21). Where resources were scarce, this inevitably meant that in order to fortify the man's role in the public domain, premium food would be diverted to him, or to children capable of bringing in income. Thus, in the Manchester slums early in this century,
dining precedence in the homes of the poor had its roots in household economics: A mother needed to exercise strict control over who got which foods and in what quantity. Father ate his fill first, to 'keep his strength up,' though naturally the cost of protein limited his intake of meats. He dined in single state or perhaps with his wife. Wage-earning youth might take the next sitting, while the younger end watched, anxious that any tidbit should not have disappeared before their turn came. (Roberts 1971:84)

There was usually little meat in the diet of the English working class during the 19th century and well into the present one (Drummond and Wilbraham


1. Overview of Trends in Dietary Variation

1958:329-30; Oliver 1895), but what little there was, was likely to be consigned to the working husband. According to Jessie Dawson, a pioneer women's health worker in the Edwardian era, this was very likely to remain true even if the wife was herself working; she concluded that "working mothers sacrificed themselves for their husbands and their children while they themselves lived in 'semi-starvation'" (quoted in H.udd 1982:xvii-xviii; Oliver 1895). Similar situations recur so often in diffeling cultures that it must be concluded that even without explicit prohibitions on the diets of women, significant dietary differences between men and women have existed routinely, though it is perhaps only in the situations of dire poverty created by industrial capitalism that the disparity has reduced the nutritional situation of women to a precarious level. It is difficult to know how much more difference pregnancy- and lactationrelated taboos could make, especially when one bears in mind that in preindustrial populations, being pregnant or breast-feeding was probably woman's general condition during her child-bearing years, so that such taboos, in fact, described not an occasional constraint on intake, but a more or less continual state. What they characterize, then, is no less than what I have been describing: a general tendency since very early tirnes for males to capture a greater share of high-protein foods. What I have suggested, moreover, is that the impetus for this has not been entirely dietary (though it certainly has had significant nutritional consequences) and that the \vay in which such esteemed foods are channeled also tends to reflect and reinforce social relations and status hierarchies. Regardless of actual practice, even the beliefs about their proper usage may play such a role.

Ecology, Economy, and Domesticated Animals in Europe
In looking, in an earlier section, at taboos in the Neotropics, I tried to demonstrate that the valuation of certain animals as food varied according to various material factors that affected the costs involved in obtaining them, relative to alternative resources. In general, despite the many obvious differences between societies that depend on hunting and those that depend on animal domestication, much the same argument applies to the latter. Indeed, in most instances there was probably little discontinuity between hunting and early domestication. (For example, the distinction between wild pigs that were hunted and pigs that were "domesticated" but allowed to forage most of the time in the forest, where they interbred with "wild" pigs, was not so very great.) Which species rose to prominence as domesticated animals in one region or another thus depended very often on circumstances not altogether dissimilar to the


I. Theoretical Overview

ones that made them favored hunting targets. One of the most important factors in this respect has always been the dietary habits of such animals themselves, a biomorphological characteristic that constrains their use Gewell 1969); even in contemporary capitalist economies, this remains an important determinant of what animals are raised where. Thus, in Europe, the distribution of such animals as pigs, cattle, and sheep has long depended upon their different feeding habits and needs-and upon the climatic conditions that favored them. Cattle and sheep, being ruminants, have been most prominent where there has been grassland, whereas swine have tended to be more important where concentrated natural feeds such as nuts, rather than bulky roughages, have been available. This basis for their distribution, in space and through time, has considerable temporal depth. From the earliest periods of animal domestication, Western Europe's extensive cover of broad-leafed deciduous forests, was chiefly occupied by swine that foraged on wild acorns and beechnuts. The distribution of cattle and sheep on the other hand, was largely determined by the absence of forest and a consequent limitation upon pig raising; it depended as well, however, upon adequate feed, which meant that, to the north, it was limited by the supply of winter fodder and, in the more arid, drought-prone, southern regions of Europe, by the difficulties of summer grazing (Pounds 1973:291). Thus, cattle tended to be most numerous in areas of northwestern Europe where the climate favored extensive natural pasture, whereas sheep, which were more hardy, assumed importance in the dry zones around the Mediterranean (ibid.: 291-92). If one bears in mind that several millennia of human activity have profoundly transformed the European landscape, so that extensive forests where pigs roamed have given way to pasture where cattle or sheep graze or have been cleared for human occupation, while, in tum, new crops have often created new conditions in which to raise swine (see Ross 1983), it is particularly noteworthy to what extent European animal domestication remains regionalized by climatic and environmental factors. Thus, as far as ruminants are concerned, cattle predominate in Europe's central latitudes, and beef consumption is low in the cold Scandinavian countries and arid southern regions such as Greece, Spain, and Portugal, where sheep far outstrip cattle (United Nations 1981: 11822, 130-32). The role of pigs is no longer defined by deciduous forest-although in some parts, such as southwestern Spain, this old association remains intact (Parsons 1962)-but by the availability of other concentrated feeds, whether crops such as maize or potatoes or waste-products of human activities. As a result, unlike grazing, but rather like milk-producing, the rearing of swine is closely identified with areas of comparatively high human density where intensive rather than extensive forms of land use are required, where farm units have evolved into small enterprises with a premium on efficiency. Thus, the highest rates of pork and bacon consumption are found in Northern and Central Europe, a ~as that have some of the highest population densities in


1. Overview of Trends in Dietary Variation

the world: countries such as Germany, Austria, Belgium, and Luxembourg (United Nations 1981: 118-22, 130-32). If it was forest that originally helped to make swine so important in the early European diet, it was the eventual contraction of European forests that caused pork to decline in dietary significance. Even in the Neolithic, when the process of deforestation began, the archaeological record indicates a notable adverse effect on the prevalence of swine (Clark 1947). The general decline of the importance of pork, however, occurred over numerous centuries during which anthropogenic grassland gradually took over from oak and beech, in the wake of land clearing for agriculture and other hunlan industries, such as building and smelting, which required timber. Thus, pigs were still of great importance well into the Middle Ages, although this often depended upon special edicts to protect acorn-producing districts, where foraging was at times a crown privilege (Steven and Carlisle 1959). The erosion of forest, however, accelerated around the 13th century, a period characterized by a dramatic rise in population and agricultural expansion (Duby 1972, 1974; 'iVhite 1964). By the fifteenth and sixteenth centuries, the distribution of pigs had been reduced largely to urban zones. There they could still thrive in the refuse in city streets, whose garbage they transformed into flesh that had become such a luxury that the rich often incorporated it into dowry payments (Forster and Ranum 1979). The decline of forest was, inevitably, not uniform. As I have already noted, there are areas in Spain where oak forests still survive-albeit through active human efforts-and continue to sustain enormous numbers of swine and to guarantee pork a special place in the diet (Parsons 1962:214-15). By the same token, certain regions were especially vulnerable to human activity and were deforested earlier and more rapidly than the rest of Western Europe. One such area was highland Scotland. The case of Scotland is enlightening, not just because it illustrates the relationship between change in dietary custom and the kind of environmental transformations that I have been discussing, but because it also provides an excellent opportunity, in the context of a complex sociopolitical system, to examine how "taboos" seem to emerge and wane in changing material circumstancesmuch as we saw in our earlier consideration of neotropical societies. Deforestation in the Scottish Highlands \vas particularly rapid, as a rugged landscape and harsh weather had limited the spread of deciduous vegetation, encouraging instead the development of a hardier pine forest that was much less favorable to foraging hogs. Archaeological evidence testifies to the fact that swine once did inhabit these upland regions (elutton-Brock 1976; Ritchie 1920), and we know that even into the 12th century Scottish highland lairds and various monkish orders raised herds of swine and ate pork (MacKenzie 1935:57). The practice does not seem to have survived the process of forest erosion, however. Agriculture, perhaps the principal factor behind deforestation, was largely


1. Theoretical Overview

confined by rain, wind, and cold to the same restricted glens and valleys where oak and beech found a limited haven. The expansion of agriculture was thus necessarily at the expense of pigs. Nor could it offer much to substitute for mast. Given the unusually harsh climate and generally poor soil conditions (Grant 1961: Handley 1953; Watson 1964), agricultural harvests were precarious and unreliable and rarely produced enough for human consumption, let alone surplus for animals. Cattle, at least, could be grazed on upland pasture and moorland, which cover more than 65 percent of the surface area of the Highlands. Few might survive the long, arduous winters (Handley 1953:69-71; Lea 1977), but at least those that did not had not taken food out of human mouths, while those that did could be restored in the spring to provide milk and cheese or sold to pay rent and enable people to buy oats with which to supplement their diet of dairy products (Grant 1961; Keltie 1875; Smout 1969). On the other hand, with the demise of deciduous forest, the pig had become "the farm animal least able to live off the country" (Darling 1955:326), and evidence suggests that although their numbers declined, those that remained acquired a powerful reputation for a rapacious omnivory that threatened everything from crops to untended infants (O'Dell 1939; Sinclair 1796:227). Such behavior prompted the levying of fines against the owners of destructive pigs and eventually led numbers of highland lairds to order their tenants "not to raise swine" (O'Dell and Walton 1962:93). It was, perhaps, out of such policies that the idea evolved of a Highland "taboo." The ecological costs of swine rearing were thus compounded by financial ones, with the result that by 1650 at least, Highlanders-who by 1578 were described by Bishop Leslie as taking "lytle plesure" in swine flesh (quoted in MacKenzie 1935: 58)-rarely raised hogs any longer and were said to regard the consumption of pork with such disdain that some observers could only attribute it to a deep-seated religious prejudice or taboo (Findlater 1802; Robertson 1799:326; Smith 1798; Salaman 1949:348; Smout 1969: 132). An English song, "The Brewer," published in London in 1731 but written some 70 to 100 years earlier, made reference to "The Jewish Scots that scorn to eat / The flesh of swine" (quoted in MacKenzie 1935:43-44). Most attempts to ascribe the Highland disaffection for pigs to religious taboo rest upon a notably arcane logic: for example, Donald MacKenzie's argument that since, among Celtic peoples, the rare group that disliked pork were the Galatians (a fact he attributes to their conversion to the ancient cult of Attis, who was slain by a boar), the Highland Scots must be descended from Galatian migrants (1935:66-68). Yet, as I have suggested, the explanation was actually far more mundane-and never that difficult to see. Indeed, in the 1730s, an English officer named Burt had written: "I own I never saw any swine among the mountains, and there is good reason for it; those people have no offal wherewith to feed them; and were they to give them other food, one single


1. Overview of Trends in Dietary Variation

sow would devour all the prOVISIons of a family" (Quoted in MacKenzie 1935:45). He went on, moreover, to describe the relative superficiality of the ideological aspect of the Highland "prejudice" against pork:
It is here a general notion that where the chief declares against pork, his followers affect to show the same dislike; but of this affectation I happened once to see an example. One of the chiefs who brought hither with him a gentleman of his own clan dined with several of us at a public house where the chief refused to eat pork, and the laird did the same; but some days afterwards the latter being invited to our mess and under no restraint, he ate it with as good an appetite as any of us. (Quoted in MacKenzie 1935:45)

The Rise and Fall of Vegetarianism irL Western Europe
Except perhaps that the decline of swine we t1t further than elsewhere in Westem Europe during this period, largely because the demise of forest was greater and more rapid and agricultural potential was in general so much less, the diminishing role of pork in the Scottish Highlands conformed to a general process of dietary transformation. But pork was not the only animal flesh to depart from the European diet as demographic growth proceeded. Indeed, by the 12th century, while pork still retained a notable position in the diet, ruminants were already being affected by increasing pressure on the land from agricultural developments. As Pounds (1973:290) has observed:
Meat and milk products formed only a very small part of the human diet in most parts of Europe, and with the exception of the pig, animals primarily served other ends than the provision of food. The ox was used mainly as a draught animal, and the paucity of references to the dairy cow, or vacca, suggests how unimportant was cow's milk in most areas. The sheep was bred more for its wool than for its food value, though ewe's milk, and the cheese made from it, were locally important.

In the following centuries, the dietary picture worsened, with pork of course declining in importance, so that in areas such as England, by the 14th century, "the staple diet of the peasantry was the fatllous English 'white meats' (milk, cheese, eggs and poultry)" (Crawford and :Broadley 1938:206). By the 16th century, increasing emphasis upon cereal production was seriously displacing animal husbandry throughout much of Europe, while, because cereal prices were still rising, according to Braudel: "There was no money for buying extras. Meat consumption diminished in the long run until about 1850" (1967: 130). By the 17th century, the emerging similarities to the dietary economy of China were quite remarkable. In China, for example, cattle were con-


I. Theoretical Overview

served for agricultural labor, and Chinese pharmacologists admonished that eating beef was unhealthy (Schafer 1977:99). Well into the present century, what beef was consumed was usually the flesh of dead farm animals (and, even then, the meat was typically sold in an urban market). In Western Europe, where cattle raising persisted, they were kept, as centuries earlier, largely as a source of rural power. There were usually enough milch cows that milk remained an important part of the diet, but what beef was eaten "often came from sick or half-dead beasts who could no longer function as work animals" (Blum 1978: 186). As in China, many physicians espoused the view that it was healthier to eat meat in moderation (which, of course, in the circumstances was probably true). With the build-up in merchant fleets, fish came to playa much more important part in the diet, and the number of meatless days in the Catholic calendar grew until, by the middle of the 17th century, they accounted for about one-half of the days in the year in England and France (Braudel 1967: 146; Wilson 1973:31). Up until the beginning of the 19th century, Western Europe was largely a non-meat-eating region as far as most of its inhabitants were concerned. Indeed, for the working class, meat remained scarce through most of the last century, with bread, potatoes, and tea the dietary staples (Drummond and Wilbraham 1958:329-30; Oddy 1970). Yet the ecological circumstances, at least, that had brought about such a deterioration in the diet were by the late 18th century beginning to be overcome by political and economic events, particularly the opening up of new lands by colonial activities. First, from newly settled regions in the Americas had come new cultigens such as maize and potatoes, which laid the basis for a new intensification of mixed agriculture. The potato in particular opened up formerly marginal lands and supplemented grain harvests-a contribution of critical importance in areas such as Scotland and Switzerland, where climatic conditions made cereal production so problematical. As Netting has written of the latter country, where potato cultivation spread during the first decade of the 19th century:
The growing of potatoes permitted the use of a wider range of land, including highaltitude, steep, and fallow plots, for food production; potatoes required little or no new investment for cultivating and processing tools; they achieved higher caloric yields per unit of land with perhaps less labor expenditure; they were more dependable than grain in the alpine zone; and they filled a nutritional need. (1981: 164)

Throughout Europe, moreover, the tuber provided the possibility of a surplus that could be invested in animals such as pigs, initiating a veritable renaissance in swine rearing. The potato, for example, made its appearance in Scotland around the middle of the 18th century (MacKay 1955:37), and its impact on the putative religious dread of pigs was dramatic. As Robert Hender-


1. Overview of Trends in Dietary Variation

son, a Scottish farmer, wrote in 1811: "About the year 1760, one would scarce have seen twenty swine in a parish throughout Dumfriesshire; but about 1770, they began to appear more plentiful, and every farmer then kept one or two." Within a few more years, a flourishing trade in cured hams had arisen, and the county of Dumfries was sending six times as many hams to London as Yorkshire was (1811: 10-12). By the end of the century, James Robertson was able to write:
Nat many years ago, the Highlanders in general had a disgust of this kind of food, without being able to give any reason for it; but their dislike to pork has greatly worn off; in so much that in a short time this loathing of swine's flesh will be accounted a singularity even among the Highlanders. (1799: 32)

England and the Political Economy of Meat
Nevertheless, it was not primarily the introduction of new cultigens that revolutionized meat production in Europe, particularly in England. The salient development was, rather, colonialist expansion, which appropriated the comestible resources as well as other raw ITlaterials of new territories abroad. (China, by contrast, having never become the same kind of mercantile expansionist state, remained largely dependent on its own food base and thus never diverged from the more or less vegetarian diet imposed by its demographic pressure for agricultural intensification.) Beginning as early as the 17th century, England had begun to achieve hegemony, by military, political, and financial means, over the pasturelands of her Celtic neighbors in Scotland and Ireland. Indeed, climate and rugged terrain were not the only constraints on agriculture that faced Highland Scots by this period, for, just as the English were conquering Jacobite forces and dismantling the feudal structure of the Scottish clans, so 'was the English market insidiously eroding the traditional subsistence economy. At first this entailed a demand for cattle; but by the second half of the 18th century, industrialization of the English textile industry was calling for new reserves of raw wool, so that flocks of sheep began to displace both Highland cattle and the people who had subsisted predominantly on the dairy products that they supplied (Ross 1983: 104-5). The Highlanders themselves were forced to leave for the lowlands or the coasts, while the wool went to the mills of Yorkshire and Scottish mutton ended up in the butcher shops and markets of London. By the same period, Ireland-which had been drawn into the orbit of England's control since the 12th century-was producing beef and pork for the use and profit of English settlers. Even more important at this date was the fact


1. Theoretical Overview

that Ireland, whose land was now largely in the hands of English colonists and crown grantees, had become a vital component in the operation of England's Caribbean colonies, to which it supplied large quantities of barrelled beef and pork (O'Donovan 1940:49). As observed in The Interest ofEngland in the Preservation ofIreland, published in 1689: "The islands and plantations of America are in a manner wholly sustained by the vast quantities of beef, pork, butter and other provisions of the product of Ireland" (quoted in O'Donovan 1940: 73). By the second half of the next century, however, the bulk of Irish livestock, meat, and dairy products had begun to shift in increasing quantities to England, largely as a result of the rise of the factory system, which drew English labor out of the countryside and put great pressure on domestic food supplies in the new manufacturing districts. There were, as a result, shortages and rising prices, which industrialists were eager to moderate for several reasons. First, they were becoming a major focus for working-class discontent and protest by the end of the 18th century (Morton 1979) and thus threatened social and economic stability. Second, increasing rivalry with other developing industrial nations in Europe compelled English factory owners to sell at competitive prices, which required holding down wages-a process that, in turn, was seen to depend to some degree on the cost of basic foods. It was these economic imperatives, as much as anything else, that lay behind England's impulse to procure Scottish and Irish meat (Stamp 1975: 240) and, in the end, to base her own industrial growth on the relegation of these peripheral regions to an ancillary role as pastoralist food reserve. By 1850 much of the meat in London's markets was "imported" from England's Celtic fringe. In that year, for example, Henry Mayhew wrote of Newgate, which he called "the greatest meat market in the world," that "one-fourth of the beef and mutton consigned to [it], is from Scotland. . . . Aberdeen is the place from which the largest supply of beef is obtained; the greatest quantity of mutton is sent from Edinburgh and Leith." Of pork, he noted, "The great supply. . . is from Ireland; from Belfast, Derry, Drogheda, Dublin, Waterford, Cork, and Limerick; and from Inverness and Invergordon in Scotland" (1850: 204-5). The effect on Scotland and Ireland was enduring. Even today the Highlands, which embrace almost half the land area of Scotland, contain only 5 percent of the country's population, and there are 10 times as many sheep as people (Grant 1961; Grieve 1972); Ireland-though three-quarters free of English political rule-remains a "cattle-farm" for its island neighbor. A sign in a typical butcher's window in England testifies to the persistent heritage of the ecological division of labor upon which modern English industrial society was based when it advertises: "Scotch beef, English lamb, Devon pork, Irish bacon." In the 18th century, the impact of English colonialism on Ireland was particularly devastating-especially in terms of the diet of the peasantry whose labor produced the commodities that found their way into English stomachs. Up until


1. Overview of Trends in Dietary Variation

the mid-17th century, the rural diet continued to consist largely of milk, cheese, grains, and some pork from swine that foraged in local forests (Lucas 1960). This pattern was disrupted by the desire of English landlords to reduce subsistence acreage and increase the land area allocated for production for export. At this date, no one can say whether it was landlord or peasant who most clearly recognized the value of the potato as an acre-economizing cropwhether landlords encouraged tenants to cultivate it in order to squeeze more people onto less land, or peasants intensified their use of this new crop as they came to recognize that it could compensate them for the reduced acreage onto which they were forced-but one way or the other, as the years went by, the more Ireland's economy was subordinated to the English market, the narrower the rural diet grew and the greater the role of the potato in it until, by the early 1800s, it was practically the sole food of the Irish peasantry (Dutton 1824: 352; Salaman 1943). (By contrast, in Switzerland" which remained economically autonomous, the consequences of adopting the potato were more unambiguously beneficial [Netting 1981: 159-69].) The result, in tum, was that their subsistence base grew increasingly vulnerable to any disruption of the potato harvest. Thus, in 1846, when a hitherto unknown blight destroyed most of the potato crop, a terrible famine devastated the rural population, causing widespread death by starvation and disease and emigration on an unprecedented scale. What was most interesting was that in the midst of this crisis, the export of Irish grain and livestock not only continued but intensified. Indeed, far from inducing any modification in English colonial rule, the Great Famine of the 1840s actually provided England with an incredible opportunity, while the peasantry was virtually prostrate, to intensify pre-famine trends in land clearance and cattle rearing. The result was that bet'Neen 1851 and 1910, land in pasturage (including grass, meadow, and clover) rose steadily from 10.7 to 13.1 million acres (O'Donovan 1940). Just from the period 1846-49 to 1870-74, the number of Irish cattle exported to England climbed from almost 202,000 to about 558,000 (ibid.:213-14). By 1880 it could be written:
Agriculture of most other kinds has been steadily dwindling down; 519,307 acres out of a total tillage area of 5,500,000 had gone out of cultivation in ten years. The wheat culture was ruined. . . . The breadth of land even under oats had declined by 320,000 acres. . . . 50.2 percent of the entire surface of the country and two-thirds of its wealth, were devoted to the raising of cattle. (Dublin Mansion House Relief Committee 1881:2)

By the 1890s, about 65 percent of Ireland's total meat production went to England, where it constituted on average about 30 percent of what English officialdom described as the "domestic" supply. Yet during the first half of the


1. Theoretical Overview

century, as an ever greater proportion of the English population, swelled by a great demographic explosion, had been drawn into industrial production, the English demand for food had already reached levels for which the food reserves of Scotland and Ireland proved increasingly inadequate. By mid-century it was already apparent that England's diet in general depended upon foods imported from an ever more diverse array of places. Thus, by 1856 George Dodd was writing in his book, The Food ofLondon:
Let the query be, whence does London obtain its butchers' meat? We find that live stock is brought to the metropolis by all the great turnpike roads, by seven railways, and by steamers. . . . We find that these animals arrive from almost every part of England, from the Lowlands and the HigWands of Scotland, from Wales and from Ireland, from Denmark and from Holland. . . . [There are also] bacon and salt meat from Ireland; hams from Yorkshire and Germany and Spain; game from all the sporting counties of the north; rabbits from the southern counties and from Ostend; poultry from half the counties in England; eggs from Ireland and France. . .. If the daily bread of the metropolis be the subject of inquiry, we must travel yet farther to trace the sources of supply; to the southern Russians around Odessa and Taganrog (unless war interrupt), the Moldavians around Galatz, the Prussians around Danzig, the Americans in the Mississippi States, to say nothing of nations nearer home, or of our own farmers. (1856: 102)

Thus, the dietary resources of England were increasingly extra-territorial, dependent not primarily upon local subsistence but on commercial and political relations with other regions around the globe. I have already alluded to the domestic political pressures that were implicated in this process-the calculated recognition that too great a rise in food prices, especially of meat and bread, would fuel social discontents and class tensions-but it is important to note as well that as England's role as an imperialist power became more and more the basis of its food supply, so political realities shifted to a wider stage, one on which it was necessary to maintain a constant protective guard over the mercantile agents of its emergent world economy. Nor should it be forgotten that foodstuffs from around the world were also a lucrative field for investment, and that the range of edible commodities that flowed into England (as, to a lesser extent, into other Western European markets) was less the result of a desire to vary the national resource base for its own sake (though in that diversity lay a certain political security) than a reflection of the profitable diversification of England's investment of finance capital. Having begun much earlier in the Celtic regions, in time this process had extended into the continental arena-where it was a powerful motive force behind the hegemonic struggles among the varied European powers-and by the late 19th century had reached as far as Australasia and the Americas. In the course of the 19th century, for example, much of Latin America had come


1. Overview of Trends in Dietary Variation

under the sway of English commercial interests, primarily in the field of textiles. By 1880 English entrepreneurs virtually controlled the guano trade and the transportation network of Peru and nitrate production in Chile, while the face value of British capital investments in Latin America as a whole had reached more than £179 million (Rippy 1959: 25-26). In the following decades, English capital was directed increasingly into the field of meat exports. Argentina is a case in point. In the 1870s, 'with its vast grasslands, it was one of the potentially most productive and lucrative regions for meat production in the Americas. But like similar frontier zones at this time, before the advent of modern transport and refrigeration facilities, its ranges were largely devoted to sheep rearing, since wool was a far less pelishable commodity. Cattle served primarily to provide meat for the Argentine consumer, while tallow and hides were generally shipped abroad. But an estilnated 45 percent of sheep and 20 percent of cattle herds were already owned by English, Scottish, and Irish colonists who had good reason to regard Europe as their ultimate market (Hanson 1938:7-10). It was the development of refrigeration that enabled Argentine beef (and, equally, that of North America during the same period [Ross 1980:202]) to be developed for export. But the impetus behind the export of beef was at least twofold: first, there was the English market, which was a significant factor underlying many innovations in food preservation during this period (Hanson 1938: 18); and, second, there was English capital, which was the pivotal element in developing the requisite infrastructure to sustain an export economy in such regions. Thus,
in the River Plate region livestock improved by the continued heavy import of British blooded animals, moved on British-owned railroads to British-equipped and financed plants, from which the finished product was shipped on British steamships to England, for a long time the only import market. (Hanson 1938: 18)

Perhaps nothing demonstrates so clearly the closely interwoven development of English diet, food importation, and rnodern English industrial capitalism in the late 19th century as the fact that one of the wealthiest families in Britain today-the Vesteys-made its original fortune, as Liverpool merchants, by dominating the Argentine meat market and going on to establish Dewhurst, Ltd., the largest chain of butcher shops in the country (New York Times 1980). To the north of Argentina, Uruguay too ,vas transformed into a supplier to the English diet, but in an altogether different way. By the 1880s over a million and a half cattle were slaughtered annually. I~eef was mainly consumed domestically, but one-third of these animals were killed only for tallow and hides, which were exported. Nearly 150,000 of them, however, ended up at the English factory at Fray Bentos on the Uruguay River, where they were rendered


1. Theoretical Overview

into "Liebig Extract of Meat"-every 110 pounds of which represented the distillation of 15 cattle (Consulate-General of Uruguay 1883:48-57). This extract, spread on bread by the English working class, became one of their chief sources of protein in the late 19th century.

The English Working-Class Diet and the Nutritional Costs of Empire
As we saw in our discussion of pre-industrial foraging and non-stratified horticultural societies, inequalities in diet emerged as a consequence of a sexual division of labor. But, in general, these were limited by the need for material reciprocity between men and women, as well as by structural constraints upon the potential for monopolization of resources. With stratified social systems, on the other hand, and especially with the evolution of class-structured systems, access to strategic resources was defined by differential power and expressed by monopolies over land or the forces of production. In Norman England in the 13th century, for example, this was clearly demonstrated by the legal concept of the "royal forest," a domain which covered about one-quarter of the surface of England and was considered quite distinct from land that the monarch held in his or her own right. In such areas habitation and all economic activities-such as foraging swine-were a source of crown revenue (Young 1979:5-6, 57). It was, in fact, the conflict between king and barons over this monopoly that led to the Magna Carta; but it was not until about 1400 that the prerogative of the crown was actually transformed into the more general privilege of the ruling class of which the monarch was the paramount figure. In 1390 a statute
ordained that no artificer, laborer, or other layman with lands or tenements valued at less than forty shillings a year nor any priest or clerk with a preferment not worth ten pounds a year should hunt deer, hares, conies, or any other "gentlemen's game" on penalty of one year in prison. The English game laws from that time marked a change from the medieval pattern and formed an alliance of king and landed gentry to protect their privileges from the lower classes. (Young 1979: 169)

In such circumstances, many of the poor could obtain sufficient food only by poaching. But the assault upon class privilege that this entailed was countered by draconian laws that regarded poaching as a criminal offense-not because it jeopardized the diet of the rich in any substantive way, but because it threatened the social order, of which dietary differences were but a reflection. In the 18th century, the relationship between diet and rank was best exemplified by the uses of venison, which, distributed as a gift or apportioned


1. Overview of Trends in Dietary Variation

among dinner guests, was probably more of a mechanism for bestowing or repaying social or political favors than it was ever a source of nutrition among the well-to-do (Thompson 1977: 158-59). In the Scottish Highlands, social access to venison became even more well established. During the second half of the 19th century, where lands from which the indigenous peasantry had been brutally cleared to make room for cattle and sheep were now apportioned among great estates (whose ownership then and now provides an excellent cross-section of the British ruling class), deer parks-eventually amounting to three million acres or 15 percent of the land area of Scotland (McEwen 1981:9; Ross 1983: 105)-were created to provide outdoor sport for the Scottish and English aristocracy. They too, in due course, were forced to cope with poaching by those few Highlanders who remained. It was in the 18th century, however, that the extent to which the parameters of diet were, in fact, also markers of class differentials in power was made most clear. In 1723 the so-called Black Act defined the criminality of poaching in the most extreme terms imaginable. According to Thompson:
The main group of offences was that of hunting, wounding or stealing red or fallow deer, and the poaching of hares, conies or fish. These were made capital if the persons offending were armed and disguised, and, in the case of deer, if the offences were committed in any of the king's forests, 'whether the offenders were armed or disguised or not. (1977: 22)

By the 19th century, with the majority of the working class now ensconced in urban districts, laws safeguarding the privileges of the landed gentry were less important than those protecting the factory owners. Thus, the political upheaval generated by the Corn Laws, which had sought to subsidize the gentry through price supports for their grain, had brought a recognition that as industry captured the leading role in the economy, it was more important to moderate the price of bread and create a more secure environment for the growth of manufacturing. But the stratification of the English diet, far from abating, intensified under the new industrial regime. The contrast in breakfast provides a useful illustration. The morning meal of the gentry frequently assumed prodigious proportions during this period. As one authority observed in 1887:
In a country house, which contains, probably, a sprinkling of good and bad appetites and digestions, breakfasts should consist of a variety to suit all tastes, viz: fish, poultry, or game, if in season; sausages, and one meat of some sort, such as mutton cutlets, or filets of beef; omelettes, and eggs served in a variety of ways; bread of both kinds, white and brown, and fancy bread of as many kinds as can conveniently be served; two or three kinds of jam, orange rnarmalade, and fruits when in season; and on the side table, cold meats such as ham, tongue, cold game, or game pie,


I. Theoretical Overview

galantines, and in winter a round of spiced beef. (Quoted in Read and Manj6n 1981:27-28)

In contrast, the breakfast of the working class usually consisted of little more than bread and butter and tea (Oliver 1895), and certainly provided few resources with which to titillate fastidious appetites. 4 Nor was their general dietary picture much better. Through most of the 19th century (and well into the 20th), it included a narrow and precarious range of foods, with bread and potatoes the dietary staples and intake of milk, meat, and fat rather low and problematical (McCabe 1974; Oddy 1970). Well-sweetened tea also came to play an increasingly important role-to such an extent that even today many working people in England refer to their evening meal, whatever it consists of, as "tea"-but, significantly, it depended more than any other item of the diet on England's colonial position. By 1880 the United Kingdom as a whole was importing over 45.5 million pounds of tea just from India (House of Commons 1871, 1888), where tea export on such a scale necessarily reduced subsistence agriculture still further in a country where colonial commerce had already so threatened food production that famines had become a recurrent threat (Bhatia 1967:31; Dutt 1940: 132). There is no doubt that as new supplies of more substantial food were secured from abroad, certain sectors of the population benefited, including skilled workers. But the English population was also on the rise, and what meat, for example, was imported does not seem to have been sufficient to preclude often severe rises in prices (McCabe 1974: 138-39), which effectively deprived many working-class families of access to such food. Thus, as Drummond and Wilbraham (1958) argue in The Englishman's Food, the opening of the twentieth century saw malnutrition more rife in England than it had been since the great dearths of medieval and Tudor times. The impoverished quality of the diet of the working class was reflected, in tum, in infant mortality rates in the major urban areas that were on a par with those found in many Third World countries today (Ross and Harris, in press) and in such levels of physical decline that, at the time of the Boer War (1899-1902), the army's rejection of recruits-who came largely from the manufacturing districts-was running at 40 percent nationally and as high as 60 percent in some areas (Howe 1972: 181). Once we have surveyed the general process of English mercantile expansion into meat-producing regions around the world, the question naturally arises: how was such meat distributed if the majority of the population seems to have consumed so little? The obvious point to be made is that the middle and upper classes claimed a disproportionate share; often what the working class was able to afford was some meat by-product, such as Liebig's extract, or a cut of meat that was "meat" in name only. (That distinction between the quality of cuts that are not merely purchased by, but even marketed to, different income strata is


1. Overview of Trends in Dietary Variation

still a fixture of distribution in our own time. Thus, in Hampstead, one of the most affluent parts of London, it is easy to find quality filets of beef; in Hackney, perhaps the poorest borough in the city, cuts that contain less meat than fat and gristle and are "as lacking in nutritional value as in aesthetic appeal, make up at least half the window displays" of most butcher shops along the principal shopping route [Harrison 1983: 27].) Such differences in diet as a whole, not just in meat, were translated into the class disparities in mortality, longevity, stature, and so on which were still conspicuous well into the 20th century-and which today are noticeable in differences in infant birth weights; 50 percent of a sample of Hackney babies were below 3,000 grams at birth, compared with 17 percent in Hampstead (Phillips 1984: 10). I noted earlier that even within working-class families, there were important contrasts in diet, so that most meat apparently was reserved for adult males, with wives skimping themselves on their behalf and children being underfed until it was time for them to go out to work, at which point they were often "fattened up" (Rudd 1982:xvii). But perhaps more important than anything else in depriving the poor of cheaper and adequate supplies of meat was the demand placed on this commodity by the army and navy, which were entrusted with the immediate responsibility for safeguarding the interests of the English ruling class and its industrial and merchant sectors overseas. Suffering unemployment and a poor diet at home, the English worker often had to risk his life overseas in the military service, in defense of a system that otherwise consigned him to misery, in order to be employed and fed. For it seems that whatever the merits of military cooking, in quantity at least it was so far superior to the civilian dietofficially it included a pound of bread and three-fourths of a pound of meat per day through most of the 19th century (Skelley 1977:63; Smurthwaite 1980:42) -that "military authorities frequently observed that the health of recruits improved even on basic army meals" (ibid.:68). It is difficult to delineate with any precision the level of demand represented by the anned services, but it seems to have been prodigious, especially in time of war-and the British army and navy were fighting somewhere in the world every year of Queen Victoria's lengthy reign (Farwell 1972: 1). Thus, during the Crimean War (1854-55), the English government frequently contracted with pork suppliers in Cork for a million pounds at a time-not a surprising quantity when just one ship, the Duke of Wellington, a giant steamer, was provisioned at the start of service with 20 tons each of salt beef and salt pork (Dodd 1856:279). Records of the British War Office's Commissariat Department provide a bit more detail, indicating that between 1813 and 1835, the department contracted for 69.6 million pounds of Irish salted beef and 77.9 million pounds of Irish salted pork (House of Commons 1836: 514). Many Irish were able to eat meat produced by their own country only by joining the army of the country


1. Theoretical Overview

that had colonized theirs, and, of course, in so doing, helped England to establish and maintain colonies elsewhere. Thus, it was not just the case that England's global commercial interests drew upon foreign resources in order to subsidize its domestic economic expansion. It was, in fact, a far more complex system than that, for, in provisioning the armed forces, its foreign trade also made the empire a self-sustaining system, with the resources from one region helping to sustain the process of acquisition or protection of markets and colonies elsewhere. When, for example, England was sending its troop ships to South Africa during the so-called Zulu War-a conflict that ended not only with the destruction of the Zulu state but with British ascendancy over the Boer Transvaal Republic (Morton 1979:485)-it was able to call upon large quantities of provisions from the United States, Canada, and Australia. According to army records, the ships sailing between February 22, 1879, and June 6, 1879, transported over 1.3 million pounds of tinned beef alone, most of it from these countries (House of Commons 1878-79: 266-67). There is no doubt, then, that the English economy was importing meat at a considerable rate; but far less of it apparently was finding its way into the domestic economy than is commonly assumed. Much of it-and it may never be possible to say what proportion-appears to have been destined to fuel the military apparatus of the English empire. What seems clear, however, is that although this considerably enhanced the economic welfare of the ruling class, its implications for the working class were far more ambiguous. While there is little doubt that the patriotic sideshows that accompanied imperial wars and conquests did much to diminish class division at home, there is no particular reason to believe that they contributed to any significant material enhancement of urban, industrial life in terms of the shelter, employment, or diet of working people.

The Economy of Feast and Famine
While England was creating its international division of labor, the United States, with its extensive forests and vast grasslands, much of them later transformed into great grain-producing regions, was largely self-sufficient. Indeed, for a long time in the late 19th century it supplied considerable quantities of meat-salted, tinned, and refrigerated-and grain and flour to England. (In 1885 these exports included 288 million pounds of bacon, 20.5 million pounds of salted beef, 94.2 million pounds of ham, and 85.4 million pounds of fresh meat URASE 1887: xxiv]; between 1875 and 1889, over 90 percent of the fresh beef imported into Britain was from the United States [Hanson 1938:70].)


1. Overview of Trends in Dietary Variation

I have described elsewhere (Ross 1980) the various interweaving strands of the process by which the United States, evolving into an industrial capitalist nation-state, was itself impelled to explore food resources beyond its immediate territory. For all the great transformations that this entailed, however, it is important, in light of earlier discussion, to note the extent to which the respective roles of cattle and pigs continue to depend on relatively persistent ecological conditions. Thus, with swine today largely supported by grain, the distribution and intensity of hog rearing is ultimately governed by the soil and climatic factors that affect grain production. In contrast, although beef cattle must be fattened for market in feed-lots, forages constitute over 70 percent of beef cattle feed, agriculture having largely reached-if not exceeded-the limits of its effective distribution (Thomas and Hopkins 1983: 13), and range and pasture still accounting for one-third of the land in the continental United States (Fowler 1961:30; Hodgson 1976:627-29). The question of whether grain is eventually employed to feed cattle or swine is itself in large measure a function of the size of the corn surplus and of various other factors-for example, the cost of fertilizer-that affect com prices. This, in tum, means that the price of corn is a cardinal factor in determining the relative supply and, hence, the price of pork and beef, which in turn is one of the major proximate determinants of consumption patterns (Ross 1980). Nonetheless, given the persisting importance of climatic, edaphic, and other environmental factors, it would be fatuous to suggest that the ecological conditions involved in patterns of diet in the United States have not been closely interwoven with the character of the economy that deploys them-an economy, moreover, that, like that of England, is not restricted to its own territorial limits. The reasons for this are far too complicated to detail in the present paper. But it must be remembered that the U.S. economy was never autonomous, that it originated as a constituent in an Atlantic mercantile economy (North 1966), and that, as a result, the development of its indigenous resources was conditioned from the start, not just by the character of those resources, but by external economic and political forces as well. These have remained a potent influence ever since and, indeed, have intensified as the world capitalist system has become more integrated over the last century through advances in transport and communication (Ross 1980). What I wish to emphasize here, however, is how the rise of U.S. monopoly capitalism introduced powerful new factors in the determination of the American diet. With the increasing centralization of the U. S. food industry, a process that dates back to the late 19th century, there came about an increasing amalgamation of political and economic power. (The most visible manifestations of this are the career trajectories of individualls who have moved back and forth between governmental offices and corporation boardrooms and the figures from major corporations who dominate important national and foreign policy-


1. Theoretical Overview

making committees [Domhoff 1970: 111-55]. An excellent example was former secretary of agriculture Earl Butz, who had previously been a director of Ralston-Purina, Stokely-Van Camp, and other companies (Finger et al. 1974: 153). One result has been that the political and military leverage of the state has been deployed to enlarge corporate markets and to expand their sources of raw materials, not necessarily because those resources are unavailable domestically, but because it may be more profitable to acquire or process them abroad. If that means that such products-whether food or manufactures-can be marketed in the United States more cheaply and more widely than home-produced varieties, it is profit rather than patriotism that has priority. (By the same token, consumers are more apt to buy cheaper products, regardless of their point of origin.) Indeed, one of the most important consequences of the increasing intensification of capitalist production in regard to food has been precisely that "modem agriculture has become, first and foremost, the production of profits and only incidentally the production of agricultural products" (Greenfield Popular Union n.d.). We have already noted this shift in its embryonic form among pre-industrial peoples drawn into capitalist relations, who have come to hunt certain animals at least as much to sell their skins as to consume their flesh. This trend has obviously advanced much further under modern industrial capitalism, where few people participate directly in actual food production and must depend instead on the profit-seeking strategies of surrogate food-producers. These, moreover, have evolved into enormous corporations controlling vast resources whose distribution transcends any specific national or cultural boundaries and whose use is dictated by many imperatives other than that of maximizing food output. As an example of how this has affected consumer options: up until the late 19th century, pigs tended to predominate over cattle in the United States because they were more efficient meat-producing animals. It was especially important, when farming was largely in the hands of modestly capitalized holdings, that 70 to 75 percent of a pig was edible, compared with 50 to 55 percent of a steer (Thompson and Kilbourn 1941:80). The giant meatprocessing companies that arose in the 1880s, however, were able to circumvent some of the risks that plagued small farmers, such as fluctuations in the price of grain, by getting into the feed-grain business. They were also, perhaps more significantly, in a position to invest considerable capital in developing profitable uses for former waste products of slaughtering. As these uses-fertilizers, pharmaceuticals, soap, shoe manufacturing-became increasingly lucrative, indeed as they became "the gold mine of the [packing] industry" (Clemen 1923:347), the fact that less of a steer was actually edible than a pig made the former animal potentially more profitable (Ross 1980). This, in tum, played a major role in shifting the base of mass-produced meat from pork to beef.


1. Overview of Trends in Dietary Variation

The dietary implications of corporate management of food resources went dramatically further, however, as the evolution of industrial capitalism shattered any vestigial boundaries between food production and other sectors of economic life and relegated food to the status of commodity rather than that of nutrient (a transformation that is often painfully obvious when one studies the ingredients of "food" sold in a modern supermarket). Today, as a result of the great acceleration of capital concentration, even such former giants of the U. S. food industry as Armour and Swift have become subsidiaries of even more powerful and more widely diversified holding companies, themselves usually embedded in a complex network of interlocking directorships of companies embracing a diverse array of economic specialties (Ross 1980). Through such financial interconnections, the activities of corporations involved in food production are necessarily circumscribed by numerous countervailing pressures that require food to be weighed in the accounts against alternative strategies for enhancing corporate profits, ranging frOJTI oil-drilling to launching telecommunications satellites. Thus, at the end of the day, when a company such as Dow Chemical produces lettuce or Greyhound (which owns Swift), meat, food producing must be regarded essentially as financially opportunistic. As such, it is inherently vulnerable to displacement at some point by more cost-effective rivals for capital investment. Whether agriculture or the quality of diet is the worse for having temporarily benefited cornpany shareholders is not on evidence an overriding consideration, least of all for the relatively small number of great banks that maintain ultimate controlling interest in such companies and therefore effectively guide the destiny of the U.S. diet (Ross 1980:211). Under such conditions, it is certainly impossible to argue that the economy is organized in such a way as to optimize food production. The implications of this situation are far-reaching. On the one hand, it has been instrumental in dramatically reducing by many millions of acres the amount of land under cultivation in the United States (Greenfield Popular Union n. d.: 291). On the other, it has transformed the quality of subsistence in many underdeveloped countries where U. S. -based multinationals have extended their activities because land and labor are cheaper and more readily controlled and where they may take practical advantage of and, indeed, shape the geopolitical strategies of the U. S. government. The classic illustrations of this process are the so-called banana republics of Central America; but today U.S. agribusiness companies range over the ~~lobe-from R. j. Reynolds-Del Monte in Mexico to Libby in Kenya (though the latter is now owned by Nestle, a Swiss conglomerate). Nonetheless, Latin America has always represented a special domain where U.S. economic interests have been particularly domineering and where more than three-fourths of U.S. agribusiness subsidiaries in the Third World currentlyoperate (Burbach and Flynn 1980: 108). Even where U. S. companies are


1. Theoretical Overview

not directly involved, it is still often the policy of U.S. aid programs, or loan programs that the United States dominates, to develop local or regional resources in a direction that has the advantage of North American business at heart (U.S. Congress 1984:96-97). As a result, between 1964 and 1973, 87 percent of government credit in Guatemala ended up financing production for export (Burbach and Flynn 1980: 104). Similarly, as a consequence of the use of investment capital from the U.S.-dominated Inter-American Development Bank, Costa Rica was encouraged to increase the cattle raising, with the result that between 1962 and 1972, beef rose as a proportion of major exports from 2.9 to 10.0 percent (DeWitt 1977: 134-35; see Chapter 22). Costa Rica has not been alone: during the 1960s and 1970s, beef production increased significantly in most Central American countries, disturbing vast areas of forest and displacing subsistence crops; rising beef exports were accompanied by declining domestic meat consumption in those countries (Myers 1981). In the end, Central American beef exports have amounted to only 1 percent of total U.S. beef consumption, mostly in the fast-food sector, but their impact has been far greater, if only in political terms, when one considers that "it is believed that ... [such imports] have done more to stem inft.ation [in the United States] ... than any other single government initiative" (ibid.: 5)-much as food imports did in 19th-century England. But unlike England's imports at that time, these imports only add to a general excess of protein and fat consumption. One must be clear: there remain sectors of the U.S. population that are undernourished. But to whatever degree cheap beef from Central America facilitates meat consumption by the many who already eat enough, it only contributes to a pattern of consumption that is, on average, in excess of nutritional requirements (Page and Friend 1978: 195; and see Chapter 9), especially among males (Rizek and Jackson 1982: 148). The ultimate explanation for this level of consumption remains obscure, though I have made tentative suggestions earlier in this paper. For the moment, however, what is more important is the manner in which the U.S. diet is now entangled with geopolitical issues that concern the quality of diet in other countries. Much as English colonialism harvested Irish pork and beef and reduced the Irish themselves to a virtually vegetarian diet, so adding unnecessary fats and protein to the average U.S. diet means depriving others of critical nutritional needs. 5 Even before they began to devote grain-producing land to cattle grazing, the countries of Central America were already suffering from an insufficiency of calories and quality protein, relying on a mixed diet of corn, beans, and rice to provide essential amino acids and energy (Frisancho, Gam, and Ascoli 1970: 1225). As the proportions of these elements in the traditional diet have been threatened by changes in land use and declining cultivation (May and McLellan 1972:211), the biological value of the diet as a whole also has been endangered and the risk of protein-energy malnutrition, already present,


1. Overview of Trends in Dietary Variation

elevated (Araya, et al., 1981). Moreover, as these countries have sought to compensate for lost subsistence production by importing staples, food prices have tended to rise, compounding the problem of securing an adequate diet among the poor and further increasing the peril of malnutrition (Burbach and Flynn 1980: 103-5).

Understanding how the circumstances emerged that created this situation has, on the face of it, taken us far from earlier discussions of diet among foraging populations. But, as I stated at the onset, ] intended to explore variation and differentiation in human diet with the aim of confronting the web of social realities in which diet is embedded. That intention hopefully has given an evolutionary continuity to the paper and shed light on son1e of the persisting forces that, for better or worse, have shaped dietary behavior at the succession of levels of societal development. For one of the more interesting observations to be drawn from the preceding discussion is that at no time in the knowable human past or present has diet been a clear-cut question of nutritional intake. It has always been complicated by a social dimension, and the social relations through which the material questions of diet must be regarded have grown increasingly variegated and influential over time. The result is that although the biomorphological or genetic constraints underlying human consumption in the broadest sense have not significantly altered, the factors interposed between them and the ultimate dietary configuration of any given human population have multiplied considerably and assumed a pre-eminent role. Even the recent or contemporary foraging or hunting-horticultural populations that contribute to attempts to reconstruct the early human economy (and, indeed, most of the living non-human primates whose "natural" behavior helps us to model early hominid behavior patterns) live within the orbit of the nationstate and the influence of market economies, and so cannot be viewed as purely "culturally" determined economies-if such entities ever existed. It is interesting, then, that those who maintain that dietary customs emanate from within a discretely defined cultural domain do not restrict their arguments to the more isolated populations, but effectively concede that the matter of cultural determination is, like language, either universal or nothing, that even the inhabitants of the contemporary industrial state must eat what they think. This argument can only achieve, at best, a superficial credibility, and does so only through its resistance to historical analysis and an implicit insistence on the relative historical impermeability of culture. I have sought to demonstrate by example how impoverished this would leave the examination of dietary custom.


I. Theoretical Overview

Another expression of a purely cultural analysis to which I have called attention is the deemphasizing of intra-population variation. One of the dangers of cultural explanation has always been its inclination to summarize human customs in a way that exaggerates the ideological, if not the behavioral, unity of social groups, rather than calling attention to the contradictions and stresses that constitute the essential dynamic of historical process. In doing so, it not only subverts our understanding of history but indeed seems to repudiate it and thus diminishes our comprehension of both the conditions that underlie the need for social and political change and the forces that must be harnessed to realize it. As I noted earlier, one of the reasons for studying human diet is precisely that it can give us an insight into the quality of human life and, in so doing, afford an opportunity to test and question facile judgments about progress and equality. To that extent, at least, it may contribute to a better and more thoughtful understanding of the course toward necessary and productive change. In that respect, anthropology has so much to offer that it would be disheartening if it did not meet the challenge.

1. Winterhalder, commenting on this particular problem, suggests that over a long enough period of research, the model of dietary breadth will reach identity with "the resources actually captured and consumed" (1981: 68). But, this is, at best, an unfalsifiable assumption-after all, one can always claim that the period of research should be just a little longer. 2. Of course, in Western capitalist countries since the industrial revolution, the employment of women outside the home has not been insignificant, but there has tended to be a very high degree of occupational segregation by sex, characterized by notable pay differentials. Moreover, even when women have worked outside the home, they have still tended to assume the major burden of domestic labor as well (Oakley 1981: 16386). 3. This is true even for most women of the upper class, where, as in the instance of the contemporary United States, class power is "based upon large corporate wealth that is looked after by male members of the intermarrying families that are its basis" (Domhoff 1970:56). 4. Variety remains a significant distinction between the diets of the rich and the poor. A recent survey of eating patterns of pregnant women in Hampstead and Hackney, respectively a well-to-do and very poor part of London, indicates that whereas the more affluent are eating as many as 32 different foods per day (often in salads), the poor consume about 12 or 13 (Phillips 1984: 10). The general lack of variety that came to characterize working-class consumption in the last century probably was one of the reasons for their affection for the new sauces and relishes (whose very lists of ingredients described the vastness of the Empire) that added a necessary piquancy to an otherwise bland and stodgy diet.


1. Overview of Trends in Dietary Variation

5. It should be noted, of course, that it is not only U. S. companies that have this effect in Latin America. As Burbach and Flynn comment, "Land that once produced blackbeans in Brazil is now being used to grow the soybeans that are turned into animal feed to fatten cattle for the Japanese market" (1980: 105).

References Cited
Araya, Hector, et a1. 1981 Nutritive Value of Basic Foods and Common Dishes of Guatamala Rural Populations. Ecology ofFood and Nutrition 11: 171-76. Basso, E. The Kalapalo Indians of Central.Brazil. New York: Holt, Rinehart and 1973 Winston. Behrens, C. 1981 Time Allocation and Meat Procurement Among the Shipibo Indians of Eastern Peru. Human Ecology 9: 189-220. Berkes, F., and C. Farkas Eastern James Bay Cree Indians: Changing Patterns of Wild Food Use 1978 and Nutrition. Ecology ofFood and Nutrition 7: 155-72. Bhatia, B. Famines in India: A Study of Sonte Aspects of the E conomic History of 1967 India (1860-1965). London: Asia Publishing House. Blum, J. The End ofthe Old Order in Rural Europe. Princeton: Princeton Univer1978 sity Press. Bolton, J. Food Taboos Among the Orang Asli in West Malaysia: A Potential Nutri1972 tional Hazard. American Journal of C linical Nutrition. 25: 789-99. Braudel, F. Capitalism and Material Llfe, 1400-1800. New York: Harper and Row. 1967 Buikstra, J. Maize: Prehistoric Choice-Contemporary Controversy; or, Maize 1983 Misunderstood. Paper read at the Wenner-Gren Foundation Symposium 94, 23-30 October, Cedar Key, Florida. Burbach, R., and P. Flynn 1980 Agribusiness in the Americas. Nevv York: Monthly Review Press. Butzer, K. 1971 Environment and Archaeology: A n Ecological Approach to Prehistory. Chicago: AIrline. Carneiro, R. 1970 The Transition from Hunting to Horticulture in the Amazon Basin. Pro-

ceedings ofthe Eighth International Congress ofAnthropological and Ethnological Sciences, pp. 244-48. Tokyo: Science Council of Japan.
1968 Slash and Burn Cultivation among the Kuikuru and Its Implications for


1. Theoretical Overview

Cultural Development in the Amazon Basin. In Man in Adaptation: The Cultural Present, Y. Cohen, ed., pp. 132-45. Chicago: Aldine. Chagnon, N.

Studying the Yanomamo. New York: Holt, Rinehart and Winston. Mate Competition, Favoring Close Kin and Village Fissioning Among the Yanomamo Indians. In Evolutionary Biology and Human Social B ehavior: An Anthropological Perspective, N. Chagnon and W. Irons, eds., pp. 86-132. North Scituate, Mass.: Duxbury Press. Chagnon, N., and R. Hames 1979 Protein Deficiency and Tribal Warfare in Amazonia: New Data. Science 203:910-13. Clark, G. 1947 Sheep and Swine in the Husbandry of Prehistoric Europe. Antiquity 21: 122-36. Clastres, P. 1972 The Guayaki. In Hunters and Gatherers Today, M. Bicchieri, ed., pp. 138-74. New York: Holt, Rinehart and Winston. Clemen, R. 1923 The American Livestock and Meat Industry. New York: Ronola Press. Clutton-Brock, ]. 1976 The Animal Resources. In The Archaeology ofAnglo-Saxon England, D. Wilson, ed., pp. 373-92. London: Methuen. Consulate-General of Uruguay 1983 The Republic of Uruguay. London: Edward Stanford. Crawford, W., and H. Broadley 1938 The People's Food. London: William Heinemann. Darling, F. 1955 West Highland Survey: An Essay in Human Ecology. Oxford: Oxford University Press. Davis, A. Women, Race and Class. London: Women's Press. 1981 DeWitt, R. The Inter-American Development Bank and Political Influence with Spe1977 cial Reference to Costa Rica. New York: Praeger. Dodd, G. 1856 The Food ofLondon. London: Longman, Brown, Green, and Longmans. Domhoff, G. W. 1970 The Higher Circles. New York: Vintage. Drummond, ]., and A. Wilbraham 1958 The Englishman's Food: A History of Five Centuries of English Diet. London: Jonathan Cape. Dublin Mansion House Relief Committee 1881 The Irish Crisis of 1879-80. Dublin: Browne and Nolan. Duby, G. 1972 Medieval Agriculture 900-1500. In The Fontana Economic History of Europe: The Middle Ages, C. Cipolla, ed., pp. 175-200. London: Fontana.

1974 1979


1. Overview of Trends in Dietary Variation

1974 Dutt, R. 1940 Dutton, H. 1824 Ember, C. 1983 Farwell, B. 1972 Feit, H. 1982

The Early Growth of the European Economy. Ithaca: Cornell University Press. India Today. London: Victor Gollancz. A Statistical and Agricultural Survey of the County of Galway. Dublin: Dublin University Press. The Relative Decline in Women's Contribution to Agriculture with Intensification. A merican A ntrhopologist 85: 285-304. Queen Victoria's Little Wars. New York: Harper and Row. The Future of Hunters Within James Bay Cree. In Politics and and R. Lee, eds., pp. 373-411.

}fisto ry

Anthropology and the in Band Societies, E. Leacock

General View of the Agriculture of the County ofPeebles. Edinburgh: Archibald Constable. Finger, B., et al. 1974 Agribusiness Gets the I)ollar. Southern Exposure 2: 150-57. Forster, R., and O. Ranum 1979 Introduction. In Food and Drink in History, R. Forster and O. Ranum, eds., pp. vii-xiii. Baltimore: Johns Hopkins Press. Fowler, S. 1961 The Marketing ofLivestock and Meat. Danville: Interstate Printers. Frisancho, R.; S. Gam; and W. Ascoli 1970 Unequal Influence of Low Dietary Intakes on Skeletal Maturation During Childhood and Adolescence. A merican Journal of Clinical Nutrition 23: 1220-27. Goldman I. 1963 The Cubeo: Indians of the Northwest Amazon. Urbana: University of Illinois Press. Graham, S. 1975 Running and Menstrual Dysfunction: Recent Medical Discoveries Provide New Insights into the Human Division of Labor by Sex. American Anthropologist 87: 878-82. Grant, I. 1961 Highland Folk Ways. London: Routledge and Kegan Paul. Greenfield Popular Union At the Crossroads: Agriculture and Agricultural Labor in the Connecticut n.d. River Valley. Turner's Falls, Mass.: Greenfield Popular Union. Grieve, R. 1972 Problems and Objectives in the Highlands and Islands. In The Remoter Rural Areas ofBritain, ]. Ashton and W. Long, eds., pp. 130-45. Edinburgh: Oliver and Boyd.

Findlater, C. 1802


I. Theoretical Overview

Gross, D. 1975 Hames, R. 1979 Handley, ]. 1953 Hanson, S. 1938 Harding, R. 1973

Protein Capture and Cultural Development in the Amazon. American

Anthropologist 77: 526-49.
A Comparison of the Efficiencies of the Shotgun and the Bow in Neotropical Forest Hunting. Human Ecology 7:219-52.

Scottish Farming in the Eighteenth Century. London: Faber and Faber. Argentine Meat and the British Market. Stanford: Stanford University
Press. Predation by a Troop of Olive Baboons (Papio anubis). American Journal ofPhysical Anthropology 38:587-92. The Predatory Baboon. Expedition 16:30-39.

1974 Harris, M. 1977 1979
Harrison, P. 1983 Hassan, F. 1975

Cannibals and Kings: The Origins of Cultures. New York: Random

Cultural Materialism: The Struggle for a Science of Culture. New York:
Random House.

Inside the Inner City. Harmondsworth: Penguin.

Determination of the Size, Density and Growth Rate of Hunting-Gathering Populations. In Population, Ecology and Social Evolution, S. Polgar, ed., pp. 27-52. The Hague: Mouton. Hawkes, K.; K. Hill; and]. O'Connell 1982 Why Hunters Gather: Optimal Foraging and the Ache of Eastern Paraguay. American Ethnologist 9:379-98. Hayden, B. 1984 Resources, Rivalry and Reproduction: The Influence of Basic Resource Characteristics on Reproductive Behavior. In Culture and Reproduction, W. Handwerker, ed., pp. 5-21. Boulder: Westview Press. Henderson, R. 1811 A Treatise on the Breeding ofSwine, and Curing ofBacon; with Hints on Agricultural Subjects. Leith: Archibald Allardice. Hill, K., and K. Hawkes 1983 Neotropical Hunting Among the Ache of Eastern Paraguay. In Adaptive Responses of Native Amazonians, R. Hames and W. Vickers, eds., pp. 139-88. New York: Academic Press. Hodgson, H. Forages, Ruminant Livestock and Food. BioScience 26: 625-30. 1976 Hoffmann, H. Money, Ecology and Acculturation Among the Shipibo of Peru. In Explo1964 rations in Cultural Anthropology, W. Goodenough, ed., pp. 259-76. New York: McGraw-Hill.


1. Overview of Trends in Dietary Variation

House of Commons 1836 Accounts and Papers: Estimates; Army; Navy; Ordnance. London: Her Majesty's Stationery Office. Accounts and Papers 50: East India. London: HMSO. 1871 1878- Accounts and Papers: Army Estimates: Army; Militia and Volunteers. London: Her Majesty's Stationery Office. 79 1888 Accounts and Papers: 77. London: H M SO. Howe, M. Man, Environment and Disease in Britain. New York: Barnes and 1972 Noble. Izaguirre, B. Historia de las Misiones Franciscanas y Narraci6n de los Progresos de la 1925 Geografia en el Oriente del Peru. Vol. 9. Lima: Imprenta Medalla. Jackson, ]. The Fish People: Linguistic Exogamy and Tukanoan Identity in N orth1983 west Amazonia. Cambridge: Cambridge University Press. Jelliffe, D., et al. 1962 The Children of Hadza Hunters. Journal ofPediatrics 60:907-13. Jewell, P. Wild Animals and Their Potential for New Domestication. In The Domes1969 tication and Exploitation of Plants and Animals, P. Ucko and G. Dimbledy, eds., pp. 101-9. Chicago: AJdine-Atherton. Journal of the Royal Agricultural Society of England (fRASE) 1887 Statistics Affecting British Agricultural Interests. fRASE, 2d ser., 23:

Katona-Apte, ]. 1975 The Relevance of Nourishment to the Reproductive Cycle of the Female in India. In Being Female: Re/Jroduction, Power and Change, D. Raphael, ed., pp. 43-53. The Hague: Mouton. Keene, A. 1981 Optimal Foraging in a Nonmarginal Environment: A Model of Prehistoric Subsistence Strategies in Michigan. In Hunter-Gatherer Foraging Strategies: Ethnographic and Archaeological Analyses, B. Winterhalder and E. A. Smith, eds., pp. 171-93. Chicago: University of Chicago Press. Keltie, ]. 1875 A History of the Scottish Highlands. Edinburgh: A. Fullarton and Co. Lea, K. 1977 A Geography ofScotland. Newton Abbot: David and Charles. Leacock, E. 1981 Myths of Male Dominance: Collected Articles on Women Cross-Culturally. New York: Monthly Review Press. 1982 Relations of Production in Band Society. In Politics and History in Band Societies, E. Leacock and R. Lee, eds., pp. 159-70. Cambridge: Cambridge University Press. Lee, R. Politics, Sexual and Non-Sexual, in an Egalitarian Society. In Politics and 1982


1. Theoretical Overview

History in Band Societies, E. Leacock and R. Lee, eds., pp. 37-59. Cambridge: Cambridge University Press.
Lindenbaum, S. 1979 Kuru Sorcery: Disease and Danger in the New Guinea Highlands. Palo Alto: Mayfield. Little, M., and G. Morren 1976 Ecology, Energetics and Human Variability. Dubuque: William C. Brown. Loveland, F. 1976 Tapirs and Manatees: Cosmological Categories and Social Process Among Rama Indians of Eastern Nicaragua. In Frontier Adaptations in Lower Central America, M. Helms and F. Loveland, eds., pp. 67-82. Philadelphia: Institute for the Study of Human Issues. Lucas, A. T. 1960 Irish Food Before the Potato. Gwerin 3(2): 8-43. McCabe, A. 1974 The Standard of Living on Merseyside, 1850-1875. In Victorian Lancashire, S. Bell, ed., pp. 127-49. Newton Abbot: David and Charles. McEwen, ]. 1981 Who Owns Scotland. Edinburgh: Polygon. MacKay, F. 1955 MacNeill of Carskey: His Estate Journal, 1703-1743. Edinburgh: M. MacDonald. MacKenzie, D. 1935 Scottish Folklore and Folk-Life: Studies in Race, Culture and Tradition. Glasgow: Blackie and Son. Maher, V. Work, Consumption, and Authority Within the Household. In Of Mar1981 riage and the Market: Women's Subordination in International Perspective, K. Young et aI., eds., pp. 69-87. London: CSE Books. Martin, Paul 1985 Prehistoric Overkill: The Global Model. In Quaternary Extinctions: A Prehistoric Revolution, P. Martin and R. Klein, eds., pp. 354-403. Tucson: The University of Arizona Press. May, ]., and D. McLellan 1972 The Ecology ofMalnutrition in Mexico and Central America. New York: Hafner Publishing. Mayhew, H. 1850 The Morning Chronicle Survey ofLabour and the Poor. London. Morton, A. L. 1979 A People's History of England. London: Lawrence and Wishart. Murphy, R., and B. Quain 1955 The Trumai Indians of Central Brazil. Monographs of the American Ethnological Society, 24. Locust Valley, New York: ]. J. Augustin. Myers, N. The Hamburger Connection: How Central America's Forests Become 1981 North America's Hamburgers. Ambio 10:3-8.


1. Overview of Trends in Dietary Variation

Neel, ]. 1977

Health and Disease in Unacculturated Amerindian Populations. In Health and Disease in Tribal Societies, !(. Elliot and]. Whelan, eds. Amsterdam: Elsevier.

Netting, R. 1981

Balancing on an Alp: Ecological Change and Continuity in a Swiss Mountain Community. Cambridge: Cambridge University Press.

Nietschmann, B. 1972a Hunting and Fishing Focus Among the Miskito Indians, Eastern Nicaragua. Human Ecology 1:41-67. 1972b Hunting and Fishing Productivity of the Miskito Indians, Eastern Nicaragua. Actas y Memorias de XX)(IX Congreso Internacional de Americanistas, vol. 4, pp. 69-88. Lima: Instituto de Estudios Peruanos. 1973 Between Land and Water. New York: Seminar Press. North, D. 1966 The Economic Growth of the United States. New York: W. W. Norton. Oakley, A. 1981 Subject Women. London: Fontana. Oddy, D. 1970 Working-Class Diets in Late Nineteenth-Century Britain. Economic History Review 23:314-23.

The Historical Geography of the Shetland Islands. Lerwick: T. and]. Manson. O'Dell, A., and K. Walton 1962 The Highlands and Islands of Scotland. London: Thomas Nelson and Sons. O'Donovan, ]. 1940 The Economic History of Livestock in Ireland. Dublin: Cork University Press. Ogbeide, O. 1974 Nutritional Hazards of Food T;lboos and Preferences in Mid-West Nigeria. American Journal of Clinical Nutrition 27: 213-16. Oliver, T. 1895 The Diet of Toil. Lancet, June 29" pp. 1629-635. Page, L., and B. Friend 1978 The Changing United States Diet. BioScience 28: 192-97. Parsons, ]. The Acorn-Hog Economy of the ()ak Woodlands of Southwestern Spain. 1962 Geographical Review 52(1961):211-35. Phillips, A. 1984 The Cost of Eating for Two. Guardian, June 10. Pounds, N. 1973 An Historical Geography of EurojJe, 450 B.C.-A.D. 1330. Cambridge: Cambridge University Press.

O'Dell, A. 1939


1. Theoretical Overview

Read, j., and M. Manj6n 1981 The Great British Breakfast. London: Michael Joseph. Reidhead, V. 1980 The Economics of Subsistence Change. In Modeling ofPrehistoric Subsistence Economies, T. K. Earle and A. L. Christensen, eds., pp. 14186. New York: Academic Press. Rippy, j. 1959 British Investments in Latin America, 1822-1949. Minneapolis: University of Minnesota Press. Ritchie, ]. 1920 The Influence of Man on Animal Life in Scotland: A Study in Faunal Evolution. Cambridge: Cambridge University Press. Rizek, R., and E. Jackson 1982 Current Food Consumption Practices and Nutritional Sources in the American Diet. In American Products in Human Nutrition, D. Beitz and R. Hansen, eds., pp. 121-62. New York: Academic Press. Roberts, R. 1971 Salford: The Classic Slum. Harmondsworth: Penguin. Robertson, j. 1799 General View of the Agriculture in the County ofPerth. Perth: Board of Agriculture. Ross, E. 1976 The Achuara Jivaro: Cultural Adaptation in the Upper Amazon. Ann Arbor: University Microfilms. 1978a Food Taboos, Diet and Hunting Strategy: The Adaptation to Animals in Amazon Cultural Ecology. Current Antrhopology 19: 1-36. 1978b The Evolution of the Amazon Peasantry. Journal of Latin American Studies 10: 193-218. Reply to Lizot. Current Anthropology 20: 151-55. 1979 Patterns of Diet and Forces of Production: An Economic and Ecological 1980 History of the Ascendancy of Beef in the United States Diet. In Beyond the Myths of Culture: Essays in Cultural Materialism, E. Ross, ed., pp. 181-225. New York: Academic Press. 1983 The Riddle of the Scottish Pig. BioScience 33:99-106. Ross, E., and M. Harris In press Death, Sex, and Fertility: Population Regulation in Pre-Industrial and Developing Societies. New York: Columbia University Press. Rowbotham, S. 1983 Dreams and Dilemmas: Collected Writings. London: Virago Press. Rudd, j. Introduction to Her People, by K. Dayus. London: Virago Press. 1982 Sahlins, M. Culture and Practical Reason. Chicago: University of Chicago Press. 1976 Salaman, R. The Influence of the Potato on the Course of Irish History. Dublin: 1943 Richview Press.


1. Overview of Trends in Dietary Variation

1949 Schafer, E. 1977

The History and Social Influence of the Potato. Cambridge: Cambridge
University Press. T'ang. In Food in Chinese Culture: Anthropological and Historical Perspectives, K. Chang, ed., pp. 85-140. New Haven: Yale University Press. Mother, Child and Nutrition. Journal of Tropical Pediatrics 1:47-53.

Sharma, D. 1955 Shorter, E. 1982 Sinclair, ]. 1796 Skelley, A. 1977 Smith, ]. 1798 Smout, T.

A History of Women's Bodies. Hamlondsworth: Penguin. The Statistical Account of Scotland, Drawn up from the Observations of the Ministers of the Different Parishes. Edinburgh: William Creech. The Victorian Army at Home. London: Croom Helm. General View of the Agriculture of the County ofArgyll. Edinburgh: Mundell and Son.

1969 A History of the Scottish People, 1560-1830. London: Collins. Smurthwaite, D. 1980 A recipe for Discontent: The Victorian Soldier's Cuisine. In Report, 1979-80, 41-54. London: National Army Museum. Speth, ]., and S. Scott 1984 The Role of Large Mammals in Late Prehistoric Horticultural Adaptations: The View from Southeastenl New Mexico. Manuscript. Speth, ]., and K. Spielmann 1983 Energy Source, Protein Metabolisrn, and Hunter-Gatherer Subsistence Strategies. Journal ofAnthropological Archaeology 2: 1-31. Stamp, D. 1975 Chisholm's Handbook of Commercial Geography. 19th edition. London: Longman. Steven, H., and A. Carlisle 1959 The Native Pinewoods of Scotland. Edinburgh: Oliver and Boyd. Steward, ]. H., and A. Metraux 1948 Tribes of the Peruvian and Ecuado rian Montana. In Handbook of South American Indians, vol. 3,]. H. Ste,vard, ed., pp. 535-656. Washington, D. C.: Smithsonian Institution. Stocks, A. Cocamilla Fishing: Patch Modification and Environmental Buffering in 1983 the Amazon Varzea. In Adaptive R'esponses of Native Amazonians, R. Hames and W. Vickers, eds., pp. 2:39-67. New York: Academic Press. Tanner, A. 1979 Bringing Home Animals: Religious Ideology and Mode of Production of the Mistassini Cree Hunters. Canada: Memorial University of Newfoundland.


1. Theoretical Overview

Primate Subsistence Patterns: Collector-Predators and Gatherer-Hunters. Journal ofHuman Evolution 4: 125-84. Thomas, G., and L. Hopkins 1983 The Ecologic Aspects of Animal Protein Production. Nutrition Today, July-August, pp. 9-29. Thompson, E. 1977 Whzgs and Hunters: The Origin of the Black Act. Harmondsworth: Penguin. Thompson, R., and G. Kilbourn 1941 The Meat Packing Industry. In The Development ofAmerican Industries: Their Economic Significance, ]. Glover and W. Cornell, eds., pp. 63-83. New York: Prentice-Hall. Trant, H. 1954 Food Taboos in East Africa. Lancet 2: 703-5. United Nations, Department of International Economic and Social Affairs 1981 1979/80 Statistical Yearbook. New York: UN. United States Congress, Office of Technology Assessment 1984 Technologies to Sustain Tropical Forest Resources. Washington, D. C.: U. S. Government Printing Office. Up de Graff, F. W. 1923 Head Hunters of the Amazon: Seven Years of Exploration and Adventures. Garden City, New York: Garden City Publishing. Watson, ]. 1964 Foreword. In The Third Statistical Account of Scotland: The County of Peebles and the County of Selkirk, ]. Bulloch and]. Urquhart, eds., pp. 7-8. Glasgow: Collins. Wellin, E. Maternal and Infant Feeding Practices in a Peruvian Village. Journal of 1955 the American Dietetic Association 31:889-94. White, L. Medieval Technology and Social Change. New York: Oxford University 1964 Press. Wilbert, ]. Survivors of Eldorado: Four Indian Cultures of South America. New 1972 York: Praeger. Wilson, C. 1973 Food and Drink in Britain. London: Constable. Winterhalder, B. 1981 Foraging Strategies in the Boreal Forest: An Analysis of Cree Hunting and Gathering. In Hunter-Gatherer Foraging Strategies: Ethnographic and Archaeological Analysis, B. Winterhalder and E. A. Smith, eds., pp. 66-98. Chicago: University of Chicago Press. Winterhalder, B., and E. A. Smith (eds.) 1981 Hunter-Gatherer Foraging Strategies: Ethnographic and Archaeological Analyses. Chicago: University of Chicago Press.

Teleki, G. 1975


1. Overview of Trends in Dietary Variation

Wrangham, R. Feeding Behaviour of Chimpanzees in Gombe National Park, Tanzania. In Primate Ecology, T. Clutton-Brock, ed., pp. 503-38. London: Academic Press. Wurtman, ]., and R. Wurtman 1983 Studies on the Appetite for Carbohydrates in Rats and Humans. Journal


ofPsychiatric Research 17: 213-21.
Yost, ]., and P. Kelley 1983 Shotguns, Blowguns, and Spears: The Analysis of Technological Efficiency. In Adaptive Responses of J\Tative Alnazonians, R. Hames and W. Vickers, eds., pp. 189-224. New York: Academic Press. Young, C. 1979 The Royal Forests ofMedieval England. Leicester: Leicester University Press.



Foodways: Historical Theoretical




food preferences and aversions is necessarily a history of divergent research strategies. Idealist approaches have emphasized the discovery and appreciation of foodways in which ernie, mental, and superstructural factors appear dominant. Materialist approaches have sought to relate foodways to biopsychological, etic, and behavioral variables. Eclectic approaches have sought a middle ground, acknowledging the general influence of material variables but allowing for a number-or even a majority--of exceptions.

Cultural Idealist Approaches
In general, cultural idealists explain variations in food preferences and aversions as a consequence of "culture" (by which they mean the learned emic and mental components of social life). This strategy has resulted in three kinds of explanatory propositions: (1) food customs are said to be the consequence of idiographic-historical continuities that regress to an unknown beginning; (2) they are the consequence of arbitrary "taste," chance, or whim; (3) they are the functional symbolic or behavioral correlates or expressions of given systems of values and beliefs. To the extent that none of these varieties of propositions invoke selection principles that can account for specific observed variations in food customs (as well as uniformities) or for the occurrence of the constraining values and beliefs, they constitute a very weak form of scientific statement or none at all.


I. Theoretical Overview

Materialist Strategies
Cultural materialist strategies are based on the assumption that biopsychological, environmental, demographic, technological, and political-economic factors exert a powerful influence on the foods that can be produced and consumed by any given human population. However difficult to measure, there are nutritional needs that must be satisfied and psycho-chemical limits of taste and toxic tolerances that must be observed. Moreover, we know from the study of prehistory, history, and ethnography that human diets have undergone a profound series of changes correlated with shifts in basic modes of production. During the later portions of the Paleolithic in both hemispheres, for example, meat appears to have been a focus of much productive effort; in the transitional phases antecedent to agriculture, a broader spectrum of foods was exploited; with the development of intensive agriculture, the essential amino acids were supplied primarily through a combination of grains and legumes; finally, the spread of industrial agriculture has led fo a renewed focus on animal foods. We also know from the study of prehistory, history, and ethnography that divergences in these broad patterns are often correlated with particular local conditions of climate, soil, fauna, and flora. The convergent and divergent trajectories of human foodways therefore invite nomothetic explanations to no lesser degree than other aspects of culture or the evolution of the earth and its biofonns. These preliminary generalizations may appear obvious and hence uninteresting, but they are the beginning, not the end, of the materialist thrust. It is reasonable to expect that more complex and interesting generalizations concerning similarities and differences in food patterns should be discoverable. To operate on the assumption that they are not discoverable in any given instance of a preference or aversion seems justified only as a paradigm of last resortthat is, one to be invoked only after the formulation, testing, and rejection of plausible nomothetic alternatives.

Idealist Approaches: Food As Thought
Starting with Levi-Strauss's (1963:89) observation that "natural species are chosen not because they are 'good to eat' but because they are 'good to think,' " it has become intellectually fashionable to give research priority to the meaning of food and to its mental, ideological, and symbolic functions. Foodways in this idealist tradition convey messages whose decipherment depends on knowing little or nothing about biochemical, nutritional, economic, or other


2. Foodways

material aspects of comestibles. To the extent that food preferences or aversions can be explained, their explanation "must be sought not in the nature of the food items" but rather in the system of signs, the fundamental semantic structures of a "people's underlying thought pattern" (Soler 1979: 129). The proposition that food is good for thought before it is good to eat has many ardent supporters: "Food has little to do with nourishment. We do not eat what we eat because it is convenient or because it is good for us or because it is practical or because it tastes good" (Welsch 1981:369). For Levi-Strauss food is a kind of language whose attributes are "good to think" because they unconsciously express fundamental themes in the human psyche. Like languages, cuisines convey messages through a small number of universal binary contrasts, namely, raw/rotten, raw/cooked, cooked/rotten. For "structural" analysis, these can be joined to form a "culinary triangle" in terms of which the "structure" of every culture's foodways can be specified (Levi-Strauss 1965). On the basis of a food's proximity to the raw, cooked, or rotten comer of the triangle, one can discern something about its social meaning. Roasted foods, for example, because of their direct contact with fire, are farthest away from rotten foods. Therefore, roasts are served to honored guests rather than boiled meat. The latter is closer to the rotten vertex, since water intercedes between the food and the flame. No empirical support has been found for the universal or even singular presence of Levi-Strauss's culinary triangle (Goody 1982:215-20). But if such a universal "structure" existed, its significance for understanding preferences, aversions, and foodways in general would be minimal, since nowhere does Levi-Strauss pose the question of why the behavioral expressions of cooking, roasting, and boiling involve particular meats, vegetables, beverages, and so on in different societies.

Food as Meaning: Mary Douglas
For Douglas, as for Levi-Strauss, the main task of anthropologists interested in foodways is the decipherment of the cryptic rnessages they contain. To explain the ancient Israelite swine taboo, one does not need to study the natural history, archaeology, ecology, and economy of svrine production, nor the nutritive role and value of pork. All the important properties of swine reside in the mental images of ancient Israelite culture as revealed in the Pentateuch. Douglas's explanation rests on three such properties: (1) swine were taxonomically anomalous and hence unclean because even though they have cloven hoofs, they do not chew the cud like "proper ungulates"; (2) swine, occasional consumers of carrion, were seen as carnivores, animals whose consumption was inappropriate for God's people; (3) the taboo on pork was a marker and inten-


I. Theoretical Overview

sifier of national identity, separating the Israelites from their neighbors (Douglas 1966, 1972). It is obvious that Douglas's inferences about Israelite mental life are essentially untestable. (How can we know what the priests of Leviticus really were thinking about beyond what is preserved in the Bible?) Moreover, even if the decoding operations are correct, there is no nomothetic output. The question of why the Israelites got it into their heads that "proper ungulates" must be cud-chewers is never broached. Given that in many cultures the pig is venerated rather than abominated, the taxonomic scheme of Leviticus is more likely to be a consequence of an aversion to swine than a cause of it. This observation applies as well to the rejection of carnivorous animals. An idiographic approach to this aversion seems wholly unjustified, since there are scores of other cultures that avoid carnivores. Indeed, Hayden (1981:398) notes that "one of the more remarkable features of hunter/gatherer subsistence is the uniformly low priority of carnivores and raptors as a human food resource." Finally, as far as Israelite social identity is concerned, pig avoidance was not an unambiguous symbol. Enemies such as the Babylonians and Phoenicians also had a swine taboo; among the Egyptians there were Osirian cults that abominated pigs (Darby, Ghalioungui, and Grivetti 1977: 195); nor is it likely that pork was consumed in significant quantities by the Israelites' pastoral neighbors. More recently Douglas (1984) has turned to deciphering patterns of food ensembles-meals and dishes. Expectations concerning what is a meal and when various kinds of meals ought to take place are cross-culturally variable. Douglas has studied the emic abstract and ideological dimensions of this subject, concentrating on the ideal patterning of main dishes and subordinate dishes, everyday meals and extraordinary meals, as expressions of social values (e. g., hospitality, ethnic solidarity, ritual) entirely independently of economic or nutritional-dietary factors. The scientific value of relentlessly dematerializing food and cooking remains obscure. According to Douglas (1984:22), this approach has "special strengths" for cultural analysis "wherever we are interested in measures of behavior that are independent of income or wealth." But why would anyone propose to study foodways in stratified societies independently of income and wealth? In challenging approaches that give priority to the semantic and symbolic aspects of foodways, materialists do not seek to deny the fact that food conveys meaning as well as nourishment. Rather, the issue is the extent to which the selection of foods to convey meaning is an autonomous process that can be understood apart from the processes that are responsible for selecting foods for nourishment; or, in more extreme forms of idealism, whether the selection of foods to convey meaning actually dominates the selection of foods for nourishment. Materialists find formulations of either genre especially unacceptable


2. Foodways

if they rule out infrastructural or political-economic causation without exploring any of the potentially relevant variables. Sahlins (1976: 171), for example, has proposed that the preference for beef in the United States expresses a culturally arbitrary designation of cattle as a symbol of wealth and virility that dominates the allocation of agricultural resources and capital. As Ross (1980) has shown, this is a completely upside-down version of how beef achieved its gastronomic pre-eminence in the United States. Contrary to SaWin's proposal that the preference for beef extends back to Indo-European ancestors, the etic preference for beef over pork in the United States is of rather recent origin. The designation of beef as a symbol of wealth, generosity, and virility did not generate the ecological, technological, demographic, and political ascendancy of the beef industry; it was the ascendancy of the industry as a result of those processes that has bestowed upon beef its special symbolic pre-eminence. It is not the magic of arbitrary symbols that restricts the composition of hamburgers to beef and only beef (and beef fat), but the beef industry's political clout in Congress and the Department of Agriculture (Harris 1986: 125). Similarly onesided and premature idealist conclusions characterize much of the treatment of South American tropical forest animal food ways, but space does not permit more than a passing reference (Kensinger and Kracke 1981; cf. Ross 1978; 1980). Materialists accept the fact that some aspects of foodways express arbitrary, aesthetic, symbolic, and ritual preferences, but in view of the evidence of the constraining effects of biology, ecology, economy, and politics on foodway (see below), extreme skepticism is warranted whenever it is alleged that arbitrary aesthetic, symbolic, and ritual aspects of culture have shaped major food complexes. The principal objection to such idealist explanations of foodways is that they tend to shut off scientifically productive inquiry. Nothing is more facile and intellectually sterile than to attribute a foodway to "culture" in the guise of a given set of ideologically determined "tastes" or values. Thus, it is scientifically inadmissible to attribute beef rejection in India to the Hindu religion without enquiring why the Hindu religion is different from Islam, Judaism, and Christianity in this regard (e.g., Simons 1982:209). Ditto for attributing Aztec state cannibalism to Aztec religion (Montellano 1978) without inquiring why cannibalism is not found in other major state religions. The acceptability of "religion" as an answer to why people behave in certain ways rather than others may reveal a deep-seated cultural commitnlent on the part of Western social scientists not to press the issue of material causality against the grain of revealed and unexaminable religious truths. 'Ib materialists it is no more acceptable to explain foodways in terms of religious precepts than it is to explain human origins in terms of creationism.


I. Theoretical Overview

While appearing to stand midway between the idealist and materialist positions (or in both at once), eclecticism actually strongly biases the interpretation of particular foodways in favor of idiographic and idealist explanations. The reason for this is that the discovery of an absence of nomothetic and material causation requires less effort than the discovery of its presence. The likelihood of explaining a particular foodway of nomothetic-materialist terms is directly related to the degree to which researchers are committed to such an explanation. This does not mean that researchers fudge their data or deliberately overlook negative evidence in order to fulfill their paradigmatic expectations. Rather, it means that the limited intellectual and material investment that anthropological and other foodways researchers are willing or able to make in a particular case rules out the possibility of an even-handed exploration of idealist and materialist options. The recent history of anthropological studies of foodways demonstrates that the manifestations of paradigmatic commitments are always more sharply contrastive in practice than in principle. For example, under eclectic auspices, mid-century anthropologists amassed a corpus of food avoidances and preferences designed to prove the futility of materialist principles. These cases were uncritically incorporated as pedagogical centerpieces in graduate education and continue to serve as a model for eclectic analysis, as we shall see in subsequent sections.

Irrational and Harmful Foodways
Cases demonstrating the ability of arbitrary and irrational cultural beliefs to create anti-economic, thoroughly useless, and harmful foodways have a special significance in the history of anthropological theory and continue to merit attention. At issue is whether foodways that are entirely useless or unambiguously harmful to material well-being occur in great numbers all over the world. Materialists of course admit that not all aspects of foodways functionally enhance human health and well-being. But that a large proportion of foodways would be materially useless and harmful-without material advantage to some group if not to all groups-clashes with the basic principles of cultural selection in materialist paradigms. Robert Lowie was a prolific collector of anti-economic and harmful foodways during the first half of this century. Lowie had a reputation for hard-nosed empiricism, but on surprisingly slight evidence he repeatedly flaunted exam-


2. Foodways

pIes of what he called the "capricious irrationality" of preferences and aversions involving domestic animals or their products. Pigs: "Throughout Melanesia the pig greatly affects social prestige without noticeably adding to mass subsistence" (Lo'wie 1942:541). "It is most important that while swine appear as domestic animals in prehistoric Egypt, not one practical purpose for keeping them has ever come to light" (1938:304). Cattle as a source ofmeat: "A Shilluk keeps hundreds of cattle, yet slaughters them so rarely that he is obliged to maintain his hunting techniques for an adequate supply of meat. . . Oxen nomlally serve no purpose at all" (1960:242). "The Zulu and other Bantu tribes of South Africa ... hardly ever slaughter their animals except on festive occasions" (1966 [orig. 1917]: 82). "But these Negroes, who ... eschew a beef diet, expend enormous effort on massaging the humps of their beasts and tvvisting their horns into grotesque shapes" (1960:242). Cattle as a source of milk: "On the other hand we have the even more astonishing fact that Eastern Asiatics, such as the Chinese, Japanese, Koreans, and Indo-Chinese, have an inveterate aversion to the use of milk" (1966 [orig. 1917]:82). "There are African tribes that churn butter only to smear it as a cosmetic on their bodies" (1938:307). Sheep and goats: "The Lango ... will rnilk neither sheep nor goats.... People from sheer acquisitiveness will accumulate vast herds that cannot be economically exploited. . . . The Pangwe keep sheep and goats not to eat them ... but mainly for the pure pleasure of possession" (1938:307). Horses: "We expend thousands for breeding race horses but have never taken up the milking of mares" (1938:306-7). Pigeons: "The Bambala take good care of their pigeons. . . yet 'they serve no utilitarian purpose'" (1938:304). Chickens: "In Burma and in its vicinity--the probable ancient center of chicken-raising-poultry are not primarily kept for utilitarian purposes, and the eggs are hardly, if ever, consumed. What the natives mainly want is to use the thigh bones of the cock for divination" (1938::303-4). Dingoes: "The Australian kept his dog, the dingo, without training it to catch game or render any service whatsoever" (1938:305).

Materialist Refutations
These extravagant claims set the stage for ecological and materialist approaches to foodways at the end of the 1950s. It is important to remember that Lowie's rejection of utilitarian components vIas absolute and that this invited


I. Theoretical Overview

refutations that merely sought to demonstrate some degree of functional utility in order to justify the materialist strategy. Ultimately, of course, the success of materialist approaches to foodways requires not merely that there be benefits rooted in infrastructure and political economy, but that these outweigh the costs. In order to advance to this more difficult and problematic assessment, however, there had to be a phase of research in which it was sufficient merely to show that the irrationalist claims were false as stated. The Melanesian pig problem was one of the first to be scrutinized and reinterpreted (Vayda, Leeds, and Smith 1961). Mass ritual slaughters of pigs were seen as both a provider of animal fats and proteins and a systematic regulator of warfare and environmental depletions (Rappaport 1967, 1984; cf. Luzbetak 1954). Criticism has been directed at the strength and effectiveness of the homeostatic functions attributed to the peace-slaughter-warfare cycle (McArthur 1977), but no one would wish to return to Lowie's view that pigs are raised in Melanesia "without noticeably adding to mass subsistence." As Morren (1977) points out, through inter-group rivalries and exchange, the Maring ritual cycle promotes increased meat production in a region conspicuously deficient in animal protein (Hornabrook 1977:55). Although there is waste, not everyone bears its burden equally, giving rise to political power based on the leverage of pig distributions and exchanges. Another focus of reinterpretation was the East African cattle complex (DesWer 1965; Schneider 1957). It was established that, contra Lowie: (1) the restriction of beef eating to ritual occasions did not necessarily mean that meat would be available in larger quantities without the ritual restrictions; (2) concentration on milk rather than meat or blood might be energetically advantageous; and (3) maintenance of large numbers of poorly fed animals can be a form of "banking" against drought and other disasters. As for Lowie's total-loss dingo: (1) dingoes are sometimes (not never) used in the hunt; (2) more often, semi-feral individuals are followed and deprived of their prey; (3) the symbiosis between dingo and hunter is enhanced by keeping them as pets and letting them loose to breed; and (4) dingoes are eaten as emergency food (Harris 1986: 186-89; Meggitt 1965). Although Lowie did not mention the sacred cattle of India, the prohibition on the slaughter and consumption of beef among Hindus is a classic example of an allegedly useless and harmful foodway that on closer scrutiny turned out to have many utilities and benefits (Harris 1966; Vaidyanathan et al. 1982). Lowie's assertions about the costliness of the Chinese aversion to milk and the irrational aversion to mare's milk in the United States were also challenged (Harris 1968: 366-70; 1985: 88-108). It goes without saying, of course, that Lowie had no inkling that the digestibility of cow's milk is linked to populationgenetic differences in lactase production among adults (see below). Although the explanation of Chinese milk aversion involves more than lactose malabsorb-


2. Foodways

tion (Harris 1977: 149-51; 1986: 130-53), past ignorance of this factor shows why premature "irrationalist" conclusions must be rejected. Finally, the low level of Lowie's resistance to inaccurate data can be seen in his assertion that there was "not one practical purpose for keeping swine" in ancient Egypt. Swine were in fact eaten in dynastic times by followers of Seth, and their services were in demand for seeding and manuring crops" (Darby et al. 1977: 171-99).

Massa White Sorghum Taboo
Anthropologists continue to make a special point of identifying what appear to be irrational and uneconomic foodways, often throwing facts and logic to the wind in order to do so. One recent example \\forth considering at some length is Igor de Garine's (1980) claim that the Massa of the Cameroon reject a variety of sorghum that would double or triple their output, solely because they wish to remain culturally distinct from their neighbors. Garine is opposed to studies in which "the material aspects of food and nutrition are dealt with excellently but no observations are made on the underlying [sic] symbolic systems." He is also opposed to studies where although "food themes are analyzed at length through traditional mythology, nothing is available about food economy and daily intake" (ibid.: 40). Garine's approach, like Lowie's, is an avowedly eclectic one that seeks to combine both points of view: "Before attelnpting to seek correlations, it is necessary to establish the facts and describe food and nutrition in a given society from as many angles as possible, and not merely give priority to some aspect of reality whether it be mythology or biomass" (1980:41). According to Garine (1979:81), the Massa, Tupuri, and IVlussey live close together within the "same natural milieu," have the same technologies, know about each other's cultures, intermarry frequently, and share a common origin. Yet their foodways are distinct. Garine attributes these differences to ideological factors: "They all hold more or less the same cards but they play them differently according to motivations which are mainly of cultural origin" (1980: 53). Nothing is plainer in Garine's own accounts, however, than that the groups do not hold the same etic, behavioral, ecolo!;cal, economic, or political cards. The Massa hold the areas closest to the river bank, which historically offered the best opportunities for cattle pasture and fishing; the Tupuri live along smaller streams and wetlands suitable for intensive agriculture but not for fishing; the Mussey have access neither to large rivers nor to wetlands. These contrastive ecological-economic arrangements did not result from cultural whim but from a process of migration and political- military confrontation over access


I. Theoretical Overview

to the most favorable habitats: "The bloodiest fights between neighboring groups have been over access to permanent water supplies, dry-season pastures, and the fertile soils that surround them" (Garine 1978:44). Arriving first, the Massa "occupied the best alluvial soils close to permanent water" (Garine 1980:49); the Tupuri were next, and the Massey last. All three groups, however, are politically subordinate ot the Fulani and to the state governments of either Chad or Cameroon, which have instituted various schemes for hydroagricultural development and compulsory cotton planting and rice cultivation. According to Garine, southern Massa formerly refused to plant "late white pricked sorghum" (5. dura), which is planted by their neighbors, the Tupuri. "He who dared plant white sorghum would die before the ripening of the crop" (Garine 1980:67). Yet late white pricked sorghum is a higher-yielding variety than the red sorghum (5. caudatum) that is the traditional cereal staple of the Massa. Garine insists that although the southern Massa possessed the hydromorphic soils required by 5. dura, they refused for "a long time" to cultivate it; had they done so, it would have "doubled the yield and considerably attenuated the harshness of the hungry season" (Garine 1979:83). The reason for this irrational behavior was that by acting "non-adaptively" from a biological point of view they were able to affirm their identity as "Massa" against the pressure of their white-sorghum-eating neighbors, the Tupuri. Given the fact that the Massa traditionally emphasized cattle raising and fishing rather than agriculture, one would like to know how the substitution of late white for early red sorghum would have affected the seasonal rhythms of the Massa subsistence pattern as compared with the subsistence patterns of the Tupuri. However, even on the evidence presented, Garine succeeds in demonstrating not the arbitrariness of cultural selection, but just the opposite. We are told that the techniques for planting late white sorghum derive from the Fulani overlords of the Massa, Tupuri, and Mussey. As a result of population increase, compulsory planting of cash crops such as peanuts, cotton, and rice, poisoning of fish stocks with pesticide run-off, and the development of a cash economy, all three groups have had to change their traditional mode of subsistence. All three have taken over the planting of white sorghum: "Today pricked white sorghum has spread to most of the groups where adequate soil is available. Conversely, traditional red sorghum appears as backward, heavy to digest and dirty" (Garine 1980:67). What this case illustrates is not "arbitraire culturelIe"-Garine's term-but selection against an aversion that significantly impeded the optimization of resource use under altered infrastructural and political-economic conditions. This is not to deny that food taboos are used by ethnic groups to intensify their social identity, but to predict that such markers will be selected against if they result in punitive economic deprivations (see below).


2. Foodways

Pregnancy and Post-Partum Avoidances
Some of the most notorious cases of allegedly irrational and hannful foodways are associated with pregnancy and post-partum food aversions. Numerous instances of foodway interdictions seem to be aimed at lowering rather than raising the nutritional status of pregnant and lactating women. Protein needs increase during pregnancy, "yet repeatedly we have found taboos, superstitions, and prohibitions that serve to eliminate or reduce potential sources of protein from the diet of the menstruating, pregnant, or lactating woman" (Wood 1979: 154). Despite the widespread concurrence by anthropologists that such practices often endanger the health of ITlother and neonates, there are few if any reliable studies that support this vei,¥point. It cannot be doubted that some pregnancy and lactation aversions reduce the consumption of nutritious foods. But this does not necessarily mean that the diet is impaired: it might even be improved if alternative foods are available or if preferred foods are added. Moreover, aversions associated with nutritionally impaired diets are not necesarily the cause of the impainnent; that is, removal of the aversion may not lead to any improvement if there are prior material restraints on the availability of the interdicted items. In a study based on interviews with 1,200 \vomen in Tamil Nadu, Ferro-Luzzi (1980a) identified over a hundred food avoidances of menstruation, pregnancy, and lactation. The information was collected by asking the women if they avoided the items on a prepared list. No attempt 'Nas made to check the correspondence between emic responses and etic observations, nor was there any comparison of the etic diet of women before they became pregnant with their diets after they became pregnant or after they gave birth. Thus, it is impossible on the basis of Ferro-Luzzi's data to say anything definitive about the effect of the taboos on nutrition and health. Nonetheless, if we suppose some kind of positive correlation between the emic aversions and etic perfonnance, Ferro-Luzzi's data suggest, contrary to her own convictions, that the taboos per se result in very few, if any, critical nutritional deprivations. For example, out of 1,067 women who were customarily non-vegetarian (who at meat, fish, or eggs), only 13 percent said that fish should be avoided during pregnancy (Ferro-L,uzzi 1980: 101-2). Some 97.5 percent of women in all categories approved of the consumption of milk and milk products during pregnancy; of the few who rejected milk products, none rejected milk. Although the total list of avoided fruits was large, the only fruits commonly said to be spumed were papayas and pineapples, neither of which is widely cultivated in Tamil Nadu. The only seed with a high consensus for avoidance was sesame, scarcely a crippling loss, given its expense and the availabili-


1. Theoretical Overview

ty of numerous substitutes. The most commonly to-be-avoided grain was Setari italica, described as a "poor man's millet which most people prefer not to eat anyway." The most to-be-avoided pulse was borregram (Dolichos bijlorus), again a rarely grown and unesteemed item. Finally, it should be emphasized that "restrictions on other cereals and pulses were extremely rare" (ibid.: 107), thereby ensuring the possibility of maintaining an adequate diet, regardless of the elicited pregnancy aversions. ("Possibility" is emphasized because dietary regimens in India are for most people, including pregnant women, at best marginal. But the question here is whether food aversions create or exacerbate this situation.) Turning to the post-partum and lactation periods, we find a more comprehensive set of proscriptions. Yet there still is no indication that faithful compliance with the list of items to be avoided would necessarily result in the lowering of nutritional standards. Even if the nursing mother actually drops the consumption of a dozen or more foods that she consumed prior to pregnancy, everything depends on how much she can consume of the optional foods that remain open for her. For the first "few days" only liquid foods must be eaten (Ferro-Luzzi 1980c: 109), but these liquids include milk, rice water, soups, and coffee (with sugar). Although "the majority of women avoided [read "said they avoided"] non-vegetarian food for at least a week" (ibid.: 110), only 6 percent of nonvegetarians said that they became pure vegetarians for one month or more (ibid.: 109). Within "a few days" bread, legumes, vegetables, and rice can be added to the liquid diet. Hence, despite the restricted list of edibles, continuity of the basic diet of rice and legumes supplemented by milk, milk products, meat, and fish is provided for. In addition, the possitility remains that special concessions in terms of the quantity or quality of permitted foods may compensate for the interdicted items (see below). Some special foods may be added, but these possibilities were not investigated (Ferro-Luzzi 1977).

Emics and Etics of Post-Partum Taboos
The all-important question concerning apparently harmful pregnancy and postpartum foodways is their effect on the daily etic consumption of nutrients. This effect cannot be measured merely by collecting all the items that women say they should not or must not eat. It can only be measured by comparing what a sample of women consumed during their "normal" state with what they consumed during their pregnancy or post-partum. Moreover, it must be kept in mind that late pregnancy and lactation are often associated with shifts in other activities as well as changes in diet (e. g., relief from arduous labor, protective


2. Foodways

and inactive "lying in"). These changes should also be considered in assessing the nutritional adequacy of pregnancy and post-partum foodways. Some information relevant to the emic/etic versions of lactation-period taboos is available from Wilson's (1980) study of Ru Mada, a fishing village on the East Coast of Malaysia. As in Tamil Nadu, the woman of Ru Mada can collectively produce a long list of items to be avoided post-partum: sour and "cold foods," all fruits except bananas and durain, all vegetables, all fried foods, several species of fish, curry, gravy, and sauces. These prohibitions are to be observed for 40 days, during which the mother's activity is restricted and she must spend much of her time on a bed raised above a warming fire. Wilson constructed a hypothetical one-day lying-in diet from interviews with 50 women and then observed the actual lying-in diet of two lactating women (one of the 9th and the other on the 20th day of confinement). Here is how the two verSIons compare:

Emic Diet
Rice (bowls) Fish (small, lean, none fried) European bread roll Egg Banana Coffee Plain biscuit Black pepper Tumeric Yeast (on rice or in water) Sugar (in coffee)

Etic _Diet
Rice (bowls and rice cakes) Fish (roast, salt, fried, stewed,

European bread roll Egg Coffee Plain biscuit Tumeric Sugar (in coffee, tea, and cakes) Tea Coconut (grated)

Soy sauce
Chilis Margarine Milo chocolate drink (fortified) Condensed milk Curry (in fish) Note first that the etic diet contains eight iteIns not mentioned in the emic diet. This is a remarkable rate of discrepancy, since only two women were observed for one day each. Second, there are three explicitly interdicted items (italicized) on the etic list (fried fish, soy sauce, curry), again an extraordinary rate in relation to the number of observations. Finally, it is clear that the addition of four


I. Theoretical Overview

prestige foods-European bread rolls, margarine, condensed milk, and Milorepresents a systematic attempt to supplement rather than deplete the lactating mother's diet. Having compared this diet with the recommended diet allowance (RDA) for lactating Malaysian women, Wilson concludes that the etic nutrient intake was "low" and "inimical to the woman's health" (1980: 73). This may be true, although it should be noted that both women were less active during their confinement, spending about two to five hours a day lying on their warming beds and giving up strenuous activities such as chopping wood that they had engaged in even during late pregnancy. No one would wish to argue that the lying-in diet was optimal. But Wilson presents no evidence that it was inferior to the diet of non-pregnant and non-lactating women as a consequence of the food aversions. In fact, one of the women gained 9.3 kg during her pregnancy, about 25 percent of which was probably depot fat (Thompson and Hytten 1977: 18). Her postpartum diet was actually the same in proteins (58.9 gm per day) as her etic pregnancy diet and slightly higher in calories (2,039 per day) and calcium. It seems likely, therefore, that her lactation diet was somewhat privileged (see Rajalakshmi and Ramakrishnan 1980) in comparison with the daily intake of non-pregnant and non-lactating women, even though it did not come up to the recommended level. Hence, it is not possible to attribute the failure of Ru Mada mothers to reach the Malaysian RDA for lactating women (even if the RDA is correct for lying-in, inactive lactating women) to the influence of the lying-in food aversions. Many factors other than food taboos may be responsible for the failure to reach the RDA, such as Ru Mada's productive capacity, per capita income, the conflicting demands on family budgets for the support of other members, especially male fishermen, and the costs and benefits of rearing additional children. Indeed, it should be stressed that there may be systemic advantages in not making an all-out effort to divert resources to each lactating woman at every birth (Harris 1985). In any event there is no evidence that the mere removal of the food taboos would result in the provision of the extra nutrients recommended for lactating women.

The Significance of Allegedly Irrational and Harmful Foodways
The refutation of cases of irrational and harmful foodways (and of irrational and harmful cultural beliefs in general) has important practical implications. If malnutrition and undernutrition are largely the result of misguided beliefs, then the indicated remedy is an educational effort aimed at introducing more salutary


2. Foodways

ideas about nutrition. If, on the other hand, the causes of malnutrition and undernutrition lie in the infrastructural and political-economic nexus, then nutrition education may serve more as a sop to conscience and a source of employment than as a contribution to improved diet (DeWalt and Pelto 1977:82). DeWalt and Pelto's study of Nopalcingo, a ~vIexican community in the state of Mexico, suggests that knowledge of modern food values is not sufficient to implement household dietary improvements: "rrhe problem, then, is not education or unwise allocation of resources but 1inancial means. . . . increasing the resources that families have to work with is likely to have a dramatic effect on nutritional adequacy" (ibid.: 91-92). To repeat, materialist strategies do not deny that a significant number of foodways may in fact be harmful. Some of these may owe their existence to mistaken notions about the relationship between diet and health. But these mistaken notions are not necessarily the result of irrational cultural premises. For example, infants and small children with diarrhea are often placed on liquid diets with reduced protein intake. Research indicates that this practice is lifethreatening (N. Scrimshaw 1981:229). If we set aside the possibility that the standard folk remedy for infant diarrhea is another form of indirect infanticide whose population-regulating effects are advantageous for juveniles and adults (S. Scrimshaw 1981; 1983), the fact remains that it is scarcely an irrational or capricious cultural phenomenon. Western-trained physicians and health workers have until recently recommended similar diets (ibid.: 230), and it has only recently been demonstrated that solid proteinacious foods are indicated for diarrhetic infants and children. Given the fallibility of Western experts, especially those operating in the fields of health, nutrition, and economic development, anthropologists have an obligation to remain highly skeptical about blanket condemnations of exotic foodways as useless or harmful. It is tempting, for example, to dismiss widespread notions about "hot" and "cold" foods as superstitions, but this dismissal would not be based on any corpus of sustained research. The little research that has been done actually suggests that in India foods classified as "hot" not only impart subjective feelings of bodily warmth and burning sensations in the eyes and urinary tract, but result in elevated levels of urinary sulfur excretion and nitrogen retention (Lindenbaum 1977: 145). It is only since 1965 (Cuatrecasas et al. 1965; Bayless and Rosenweig 1966) that Western physicians and health experts have come to the realization that the rejection of cows' milk in adulthood is not an irrational cultural "taste prejudice" of backward peoples, but a physiologically justified aversion based on the genetically determined enzymatic activity normal to most of humankind. Given the recently acquired knowledge about rapid fall-off with age of lactase production among most human populations, the complete withdrawal of animal


I. Theoretical Overview

milk from children suffering from diarrhea (one of the symptoms of lactase deficiency) in India (Lindenbaum 1977: 146) cannot be dismissed as an altogether harmful practice without further research.

Toward a Theory of Foodways: Emics and Etics
We are now in a position to generalize some of the critical issues which must be dealt with in the development of a comprehensive theory of food preferences and avoidances under materialist auspices. To begin with, due attention must be paid to the possibility of differences between etic and emic versions of foodways. For example, it is generally supposed that families in the United States eat definite meals together. Videotape studies of etic behavior shows that in some U.S. urban domiciles parents may eat not at a table but on their bed or watching television, while children eat in the living room or grab snacks in the kitchen (DeHavenon 1977). Similarly, Harrison (1975) has shown that U.S. households tend to give grossly unreliable estimates of their beer consumption. An important issue in the sacred cow controversy is the discrepancy between the etic rate of bovicide as revealed by lopsided sex ratios and the ideology that insists that unwanted cattle are kept alive (Vaidyanathan, Nair, and Harris 1982). The fact that a rule exists for avoiding a particular species does not mean that the rule is uniformly obeyed. Many interdicted foods are actually consumed surreptitiously: rat-eating castes in India, for example, deny that they do so in order to escape the onus of pollution. Old people in the United States deny that they eat pet food. Jews and Moslems consume pork in restaurants out of sight of family and friends. Conformity to a foodway may depend on circumstances not explicit in the rule. Many interdicted species, such as eagles, buzzards, snakes, and dogs, are consumed during hungry seasons, droughts, and other emergencies. If these species are scarce and if they are poor sources of energy and other nutritional benefits, it would be their regular consumption rather than their interdiction that would weigh against materialist strategies.

Foodways and Status Privileges
Within-group adherence to foodway aversions and preferences may be a specific function of status position. This means that benefits and costs may be distributed unevenly among age cohorts, between males and females, and during


2. Foodways

the phases of the life cycle (e. g., pregnancy, lactation, marriage, and widowhood). No materialist principles predict that everyone benefits or suffers equally from foodways (or other cultural beliefs and practices). Proscription of a food for one status group may be the precondition, under a given set of infrastructural and political-economic arrangernents, for its consumption by others. In highland New Guinea and throughout South Asia, men protect their access to scarce choice animal foods by interdicting consumption by women (Lindenbaum 1977; 1979: 133-37). Among the Amazon Tapirape only senior adult males consume deer meat. Wagley (1977: 67) sees this and other Tapirape animal taboos as "decidedly dysfunctional." Yet the deer meat taboo is not clearly dysfunctional from the point of view of the senior males. Moreover, the question of how much deer meat there would be for anybody if everyone consumed it equally remains unanswered. Recent quantitative studies of Amazonian aboriginal subsistence point up the extreme vulnerability of tropical mammalian species to depletions (Baksh 1982; (;ood 1982 and Chapter 16; Harris 1984; Paolisso and Sackett 1982; Sponsel 1983). Many preferences and avoidances associated with gender hierarchies are best regarded as forms of rationing (Gross and Underwood 1971; Harris 1985; Maher 1981; Miller 1981). Membership in various types of egalitarian and stratified groups may bestow differential rights and obligations with respect to foodways. Apparently arbitrary specialists often reflect a systemic division of labor, an apportionment of risks, and an equal or unequal distribution of nutritional benefits. The oft-noted specialized preferences and aversions of sibs and lineages, and especially of castes and ethnic groups, suggest an underlying allocation of environmental and economic niches. So-called totemic taboos may reduce the number of people competing for eponymous game animals and plants. India is especially well-endowed with segmentary specialties based on an apparently irrational interdiction of certain species of animals and crops (Sopher 1980: 196). In Kerala, for example, there are Hindu castes of low rank that specialize in the production and consumption of fish. This may protect them from competition from higher-ranking castes for whom fish and fishing are proscribed (Klausen 1968). Similarly, the defilernent that animal slaughter and beef consumption confer on Hindus is the precondition for the economic success of Moslem butchers. In turn, the avoidance of pork by Hindus and Moslems confers a monopoly on pig raising by Christians and untouchables. Similar ethnic and caste niches characterize much of Mrican social structure and are clearly evident, though not discussed, in Garine's account of the Massa, Tupuri, and Mussey. None of this should be taken to mean that segmentary foodways necessarily moderate exploitative relationships in stratified societies. On the contrary, the balance sheet of sumptuary privileges, as in the case of the restriction of alcoholic beverages to the nobilities of Polynesia and Inca Peru, may be entirely


1. Theoretical Overview

lopsided. In general, ruling classes enjoy access to foodways that are taboo to commoners or that are too expensive for commoners to enjoy. Vegetarianism has entirely different dietary implications for upper-caste Hindus who can afford to consume copious quantities of milk and milk products and for lower-caste Hindus (see Chapter 18). Similarly, upper-class Moslems may have ready access to alcohol or other drugs while enjoying the social and economic benefits that flow from sobriety among the masses.

Use and Non-Use for Optimization
There are no materialist principles that lead to the prediction that all edible plants and animals will be used. A considerable amount of non-use is essential for optimization-selection models. Such models assume that some edible species are more useful than others; therefore, some potentially useful species will be neglected in order to optimize the cost/benefits of production and the maintenance of nutritional standards. The fact, for example, that the !Kung recognize 105 edible plant species but derive 75 percent of their plant food calories from only 14 of them (Lee 1979: 159) confonns to optimization models. Similarly, the fact that the! Kung consume only 80 out of 262 animal species (ibid.: 226) suggests optimization, not waste. Obviously the place to begin to understand the list of frequently used, as opposed to seldom or never used, species is the measurement of their abundance, their seasonality, the time and energy costs of "harvesting" them, and their nutritional value (ibid.: 168). "Tastiness," a criterion mentioned by Lee, is also a viable index of behavioral cost/benefits, provided that the standards of taste derive from biopsychologically detennined preferences and aversions (possibly the preference for sweet foods over bitter ones, for example-see below). If "taste" simply means emic preference or avoidance, however, we are obviously back in the quagmire of cultural arbitrariness that is the first resort for idealists but the last for materialists. Optimal foraging models predict that plant and animal species whose exploitation lowers the average calorie cost/benefit ratio of foraging activities will not be gathered or hunted (Hames and Vickers 1982; Hawkes, Hill, and O'Connell 1982; Winterhalder and Smith 1981; and see Chapter 13). Energetic efficiency can be maximized by concentrating on species that rank high on the optimal diet list. Alternative dietary currencies should also be considered in calculating efficiency ratios-protein, fat, fiber, vitamins, minerals, and so on. These have yet to be incorporated into mathematical models, but their effects on aversions and preferences can scarcely be ignored. The quest for fats and proteins, for example, frequently takes precedence over the quest for calories, even though hunting is a much more calorie-expensive activity (Good 1982; Gross 1982; Lee 1979:262; Speth 1983).


2. Foodways

From a slightly different angle, it is clear that the broader the range of exploited dietary resources, the greater the opportunity for problems of scheduling to occur (Flannery 1968). The need to prioritize subsistence efforts can result in the neglect of relatively abundant species. The Nootka, for example, ignored runs of pilchard fish because there were other, more important seasonal economic tasks to be performed (Hayden 1981:398). Notions of conflicting priorities, scheduling, and maximization underlie the argument that economic and ecological cost/benefits are responsible for tabooing the flesh of such specialized domestic animals as the pig, horse, camel, dog, and cat (Harris 1986). Ross's (1978) explanation of why hunting activity in some parts of aboriginal Amazonia focuses on small accessible animals such as monkeys and birds to the neglect of larger, more furtive semi-aquatic species such as tapir, capybara, and deer is also predicated on the need to prioritize effort as well as on the cost/benefits per species. Optimization principles also provide a likely solution to the widely noted avoidance of raptors and carnivores. These creatures not only tend to be solitary, hard to find, and dangerous when found, but they are notably deficient in fat, which is much sought after by hunter-gatherers and village horticulturalists. Optimization "currencies" cannot be restricted to nutritional costs/benefits, since many edible plants and animals have non-nutritive uses. The costs of avoiding the consumption of domesticated species such as cats, dogs, horses, and cattle may be systemically overridden by their multiple, non-nutritive, specialized, technological functions: cats for mousing, dogs for lookouts, herding, hunting, and scavenging; horses for riding, traction, milking; cattle for traction, milking, manuring. In other words, under conditions related to the entire infrastructural and political-economic conjunction, an animal species may be too valuable alive for it to be raised and slaughtered for food. Carroll (n. d.) has found, for example, that North American Indian groups who had access to large game animals such as buffalo or deer seldorn reared dogs for food consumption but exploited them instead for their contribution to the hunt. The widespread prohibition on the slaughter of human beings for food is founded on the benefit side on the fact that in state societies, at least, people are exceedingly more useful alive than dead; and on the cost side on the fact that people are the most expensive and dangerous of all species in either their domesticated or their "wild" form (i. e., as strangers and aliens). Optimization principles further lead to the prediction that among hunter/gatherers' the larger the percentage of potentially edible species in a given habitat that are actively exploited, the lower the overall efficiency of food production is likely to be. This follows not only from optimal foraging theory but also from the principles of ecological food chains. Band and village peoples whose diets stretch all the way from earthworms, seed, acorns, and clay to shellfish, snails, and cormorants are likely to be nutritionally vulnerable. They would be much better off energetically eating closer to the top of the chain


1. Theoretical Overview

rather than at the bottom-sea mammals and fish rather than molluscs; ungulates rather than seeds; suidae rather than acorns. As an index of "hard times" (Cohen 1977), extreme breadth of diet seems to have relevance to the development of prehistoric "broad-spectrum revolutions" (Flannery 1968 and see Chapter 10), which set the stage for the shift to agriculture in the Old and New Worlds. According to Binford (1983: 212), the phenomenon should be called a "broad-spectrum depression," not a "revolution":
As hunters and gatherers were packed in a region, they were forced to move down the animal body-size chain, to exploit a wider variety of species, to make use of increasing numbers of smaller and smaller food packages, to compensate for the more specialized (and no longer viable) strategies they employed as spatially unfettered hunters.

The extent to which this same principle can be applied to agricultural and state-level societies remains unexplored. It seems applicable in the case of the Aztecs (Harris 1978; 1986:225-34), but the effect may be masked in other states by the characteristic flourishing of haute cuisine, which leaves no edible unturned in its search for ever more exotic concoctions (Goody 1982). Closely packed peasant populations theoretically ought to be receptive to a wide variety of food options and to be relatively free of general taboos (as distinguished from status and life cycle taboos). The generalization seems to hold for the Chinese peasantry, who are notorious for making "judicious use of every kind of edible vegetable and insect as well as offal" (quoted in Chang 1977: 13). India, with its proliferation of caste and ethnic taboos, seems to be the glaring exception. Yet, in a sense, the Indian peasantry and lower classes are also notable for their willingness to eat almost anything that their habitat and technology can provide. The broad spectrum is there from insects to carrion, but the various parts of the spectrum have been allocated to different segments of the population. As far as diversity of daily fare is concerned, China is (or was until recently) probably as restricted as India. It is reported that coolies in 19th-century North China ate "sweet potatoes three times a day, every day, all through the year with small amounts of salted turnip, bean curd, and pickled beans" (Spence 1977: 267). A more favorable view of the diet of a majority of northern Chinese characterizes it as "monotonous" (Hsu and Hsu 1977:315). In advanced industrial societies, per capita income is higher and hence daily menus are more diversified. But, paradoxically, among lower- and middle-class post-peasants, the spectrum of exploited foods may be narrower than among densely populated agro-managerial societies. The United States, with its rejection of mutton, goat, donkey, horse, dog, and insects, is a conspicuous example. A plausible explanation for gastronomic provinciality lies in the availability of energetically more efficient and abundant species such as cattle, swine, and


2. Foodways

poultry (Harris 1986). The absence of famines may also be relevant. The collective foodways experience of the Chinese, for example, has been dominated by prolonged periods of malnutrition and famine during which insects were luxuries and daily fare consisted of ground leaves, sawdust, thistles, cotton seeds, peanut shells, and ground pumice (Spence 1977:261). As long as daily U.S. foodways provide abundant, reliable, and relatively cheap alternative sources of animal fats and proteins, the list of rejected animals does not constitute a challenge to optimization principles, especially in view of the important non-food functions fulfilled by domesticated species. Similar considerations seem relevant to the rejection or neglect of certain parts of animals: head, hoofs, spleen, entrails, and so on, which have alternative commercial utilities. Very little of the carcasses of cattle or swine is actually wasted by modem meat-packing industries.

Short-Term versus Long-Term Optirnizations
An important qualification of optimization theories of foodways must be made in relation to collective wisdom embodied in rituals and symbols that obstruct the use of species vulnerable to depletion and extinction. If a human group has witnessed the depletion of a species or the degrading of a habitat, it is entirely plausible that an appropriate ambivalent or negative attitude would survive the passage of several generations and persist even after the endangered species or habitat had been restored to exploitable levels. If these attitudes and their corresponding practices result in a dampening of exploitation rates, they will tend to be propagated even if short-run cost/benefits favor more intensive use. The suggestion that some tropical forest hunting taboos are the forerunners of modem environmental conservation movements (McDonald 1977; Ross 1978), therefore, does not pose an unrealistic systemic "tracking" problem for pre-industrial peoples. If one assumes that there are always etic violations and modification of taboos, the net result of re stricting use will be dampening of production and the maintenance of sustainable "harvests" as opposed to periodic or permanent crashes. The ritual restrictions on East African cattle slaughter appear to operate with a similar logic: by confining slaughter to ritual occasions, they dampen the tendency to intensify slaughter and consulnption beyond sustainable yields. Similarly, the taboo on the slaughter of cattle and the consumption of beef in India does not prevent slaughter and consuITlption by Moslems, Christians, and outcasts, but the dampening effect of the restrictions ensures a steady supply of cheap beef in India's urban markets as well as cheap traction animals for the countryside.


1. Theoretical Overview

Toward a Theory of Food Taboos
Many if not most food aversions have a secular quality to them in the sense that consumption of the interdicted item is said to bring misfortune or ill-health through some emically ordinary process. The term "taboo" might usefully be reserved for aversions of a more sacred sort-namely, those which are backed up by elaborated religious notions of obedience to the will of a deity. A recurrent criticism of optimization models holds that if aversions are initiated by adverse cost/benefits, then there is no need for the elaboration of religious proscriptions. Self-interest should be sufficient to maintain the aversive behavior, and if there is any need for an emic rule at all, it would be more effective to couch it in terms of utility rather than sanctity. The need and occasion for sacred interdictions, however, may arise from several circumstances. First of all, there is the problem of ambiguity and ambivalence. The very notion that both costs and benefits must be taken into account suggests that the utility or non-utility of an item is a matter of degree and that there will normally be arguments both for and against eating it. In this light, one can understand why the pig, so highly productive in river bank and forest settings, would recurrently become the focus of intensely ambivalent feelings and hence of taboos in a semi-arid habitat. Similarly, eggs, which are a frequent focus of taboos, present the difficult problem of whether it is better to eat the egg now or the chicken later. Snakes, some being poisonous, others not, also invite sacred interdictions. Taboos are culturally selected for their ability to overcome the ambiguity and ambivalence generated by the complex systemic interrelationships in which foodways are embedded. The voice of God dispels doubts. If a taboo is associated with higher levels of collective material well-being, it will be maintained. If not, the taboo will be withdrawn or reversed. Given the scientific community's doubts and confusions about the relationships among diet, health, and development, it is easy to understand why foodways have been such a prolific source of taboos. Doubts and ambiguities and, hence, taboos are especially likely to arise when the systemic cost/benefits of an item shift from favorable to adverse as a consequence of ecological and infrastructural changes (Harris 1979: 192-95, 248-53). A slightly different scenario envisages the elaboration of taboos as a means of resolving more overt conflicting interests and segmentary differences with respect to the exploitation of resources. Caste and ethnic foodway specialties in India are made more stable by their elaboration in Hindu theology. This does not mean, of course, that taboos benefit every segment equally (Sopher 1980: 196). Finally, there is no reason to suppose that all foodway taboos are selected for


2. Foodways

their contribution to the favorable resolution of ambiguous cost/benefits. As the proscriptions in Leviticus demonstrate, some foodway taboos may be elaborated precisely because the abominated iterns possess no significant utility or are altogether unavailable in a people's habitat. The theological justifications for such taboos may have no specific relationship to infrastructural or politicaleconomic nomothetic processes and may best be examined under other paradigmatic auspices. As previously indicated, the taboos that pose a puzzle for cultural materialist principles are those that result in a significant deterioration of cost/benefits to all segments of a society. This principle in no way leads to the denial of non-infrastructural or political-economic cost-benefit functions for foodway taboos as icons in theology or in the intensification of social identity. What is predicted is that religions that persist in tabooing systemically valuable and irreplacable foods will be selected against.

Biopsychological Determinants
It is clear that any thoroughgoing optimization model implies that there are biopsychological needs and drives that food\vays satisfy with varying degrees of efficiency. Economizing of costs implies at the very least a genetically determined propensity to conserve energy. More specific biogenetic factors shape the need for essential nutrients and undoubtedly endow human beings with certain food preferences and aversions at birth. The problem of identifying and measuring these genetically determined constraints on foodways, however, is surrounded by all of the classic difficulties of nature/nurture distinctions. Despite a vast amount of research, RI)As remain "prudent" estimates subject to wide individual and population differences of both biological and cultural origin (see Chapters 6 and 9). Research aimed at determining the relative strength of the biological and cultural contributions to the origin and maintenance of specific aversions and preferences is urgently needed as a foundation for materialist explanations of both constant and variable foodways. There is some evidence, for example, that the widespread preference for animal fats and proteins displayed by the majority of known societies has a genetic component (Abrams 1979, 1980; Hamilton and Busse 1978; see also Chapters 4 and 8). Meat is known to stimulate distinctive behavioral responses in other primates. Chimpanzees, for example, are omnivores who invest considerable effort to get a small caloric reward in the pursuit of prey species-principally juvenile monkeys and galagos. While hunting, distributing, and consuming meat, chimpanzees display distinctive social behavior, such as a higher than normal amount of requests, excitement, sharing, and social control (Goodall 1979; Teleki 1981:331). The spe-


1. Theoretical Overview

cial importance of meat in the establishment of social responses strongly suggests that meat has special nutritional significance for chimpanzees and that their meat-eating behavior expresses an instinctive recognition of that significance, elaborated and modified by social traditions and individual experience. Conjointly, predation against and consumption of vertebrate animals may in itself be locked into a genetically programmed set of responses that impart a special psychosocial significance to meat. In this connection, it should be noted that chimpanzees ignore carrion and that only actively hunted or fresWy killed vertebrates elicit intense social behavior (Teleki 1981:328-29; but see Hasegawa et al. 1983). It cannot be mere coincidence that meat and other animal products also playa special role in the foodways of most human populations, particularly as a focus of exchange, redistribution, and social cohesion (Good 1982; Harris 1986: 19-46), but this does not necessarily mean that genetic coding accounts for all or any of the cross-species similarities. In the human case, the preferential status accorded meat (especially fatty meat) and animal foods may simply reflect purely traditional cultural codings that favor meat and animal products as cost-efficient packages of calories and essential nutrients. The favored status of meat and meat products could be propagated and preserved by reinforcement in the form of the extra health, vigor, and body size conferred by high levels of meat consumption, without any genetic program for meat preference. The case for certain genetically controlled innate human taste preferences rests on firmer ground. Neonates prior to any feeding experience exhibit distinct facial responses to sweet and bitter substances: a marked relaxation of the face versus open-mouthed grimacing with a flat, protruded tongue. Similarly distinctive facial reactions to sweet and bitter are found in other mammalian species and among culturally differentiated human adults, and can reasonably be interpreted as brain-stem-controlled communicators of preference and aversion (Steiner 1979:251-52). Presumably the innate preference for sweet-tasting substances is a bioevolutionary adaptation based on the fact that sweetness is characteristic of high-energy substances and hence is a predictor of nutritive value (Rozin 1982:228). Similarly, the innate aversion to bitter substances suggests a bioevolutionary adaptation that protects against the ingestion of harmful and poisonous substances. However, even in these instances, cultural programming can override innate tendencies, as can be seen in the early morning coffee taken without sugar or cream and the "black drink" (flex vomitoria) ingested daily by southeast Woodlands and Amazonian Amerindian groups for its narcotic and emetic effects (Hudson 1979). A limited number of foodways may best be understood as examples of recent (i.e., post-Pleistocene) gene-culture co-evolution. Among Europeans, and possibly other populations, the ability of adults to produce lactase, the enzyme necessary for digesting lactose, appears to have evolved during the spread of


2. Foodways

an early neolithic dairying complex. Because of the clothing they wore and their cloudy days, long nights, and long winters, early neolithic peoples of Europe were prone to vitamin D deficiency. Vitamin D is essential for the absorption of calcium and the prevention of rickets and osteomalacia. Lactase promotes the absorption of calcium and hence can playa role in compensating for vitamin Ddeficient diets, resulting in the selection of the gene for lactase sufficiency in adulthood (Simoons 1982:215). This scenario explains why Europeans have a preference for, and Chinese an aversion to, fresh liquid milk. But it does not explain why the Chinese have never taken up dairying. Aside from the beneficial effect on calcium absorption, the nutritional benefits of milk and milk products do not depend on the ability to drink fresh fluid milk. In cheese and fermented products, bacteria do the work of lactase and break lactose down into simple sugars. By exploiting such products, India, despite a rate of lactose malabsorption similar to that of China, has become a major dairying center. The explanation for the Chinese aversion to milk must return therefore to considerations of the geographical, ecological, and political-economic features that have molded basic Chinese subsistence patterns in contrast to those of other Eurasian states (Harris 1986: 130-53). Culture-gene co-evolution also seems to be involved in shaping a preference for fava beans (Vicia faba) in certain regions of Europe, Africa, and Asia. Why fava beans are eaten at all is a puzzle, since it is well known that ingestion of the beans produces a potentially fatal hemolytic shock in some members of favaeating populations. Those at risk carry the gene for glucose-6-phosphate dehydrogenase (G6PD) deficiency. As many as 5 to 30 percent of the affected populations carry the gene. A possible explanation is that there are strong oxidant chemicals present in fava that increase resistance to malaria, both in normals and in female heterozygote carriers of G6PD. The benefits of this resistance outweigh the costs and thereby maintain the cultural preference for planting and consuming the beans as well as contributing to the preservation of a balanced genetic polymorphism (Katz 1979; 1982: 176-78; and see Chapter 5). A final complication of the study of foodway selection arises from the ingestion of harmful substances such as caffeine, nicotine, alcohol, cocaine, and other psychoactive chemicals. The biopsychology of these narcotics and stimulants obviously operates in conformity with optimizations that can override and cancel nonnal nutritional and energetic cost/benefits. Their presence or absence and their degree of use is clearly subject to cultural mediation, but the possibility of genetic population differences in the appeal and effect of various psychoactive substances should not be ruled out. Many additional gene-brain-body-nature-culture interactions relevant to foodways undoubtedly remain to be discovered. There is new evidence, for example, of direct linkages between dietary amino acids such as tryptophan and tyrosine and the synthesis of neurotransmitters such as serotonin, dopamine,


I. Theoretical Overview

and norepinephrine with consequent alteration of motor activity, mood, problem-solving abilities, sleep patterns, and sensitivity to pain (Kolata 1983). The mix of factors responsible for the distribution of preferences for salt and spices such as chili and malabar pepper, thyme, marjoram, oregano, and tarragon promises to be equally complex. Reported effects of chili (Rozin and Schiller 1980) and malabar pepper include increased salivation, flavor enhancement, sense of satiety, and reduced flatulence from diets heavy in beans and pulses. Spicing as a form of food preservation alternative to salting should not be excluded, especially in view of the health hazards associated with too much dietary salt. Thyme, marjoram, oregano, and tarragon contain aromatic deoxidants whose chemical composition is similar to the industrial preservatives BHA and B H T. Salt preferences may vary with the amount of meat and other dietary sources of sodium (cf. Denton 1983). Very little work has been done on the possibility that genetic polymorphisms and direct behavior-modifying effects may also be involved in all of the above-mentioned preferences.

The Political Economy of Foodways
Control over the production and consumption of food has been a source of power and wealth from prehistoric times to the present. Since the onset of European mercantile and capitalist expansion in the 15th century, rapid changes in world foodways have resulted from the struggle for political hegemony and the search for profit. To tell the story of the transfer of animal and plant domesticants to and from the Eastern and Western hemispheres is to tell the recent history of the globe (Mintz 1979; Ross n. d.; Wolf 1983). The effects, both beneficial and harmful, are impossible to summarize. But one trend in particular-namely, the increasing role of profit in the development of mass preferences and aversions-should be noted. Increasingly the solution to the puzzle of why certain foodways develop and persist must be found in a combination of biopsychological, infrastructural, political-economic, and profitability factors. The use of tobacco, for example, is in part attributable to an innate vulnerability to nicotine addiction, in part to the toleration of governments that allow or even encourage the nicotine habit to develop while supressing the use of other addictive drugs, and in part to the huge profits made by tobacco companies, which in tum finance scientifically designed advertising campaigns and other promotional activities. A similar set of factors must be considered in explaining the popularity of coffee, tea, chocolate, alcohol, and "soft" drinks. All of these beverages exploit innate vulnerabilities to sugar and/or addictive and/or psychoactive substances, principally caffeine and alcohol; they are protected and


2. Foodways

subsidized by governments; they are a source of profit for national and multinational corporations; and they are relentlessly promoted and advertised. The vulnerability of modem industrial populations to foodways that lead to obesity and cardiovascular disorders also arises from a similar set of causes (Turner 1981). For the first time in history or prehistory, many nations no longer confront hungry seasons or periodic famine. Even the poorer classes have access to quantities of starches, sugars, and fats sufficient to make them obese at their required levels of daily activity and calorie expenditure. With this phenomenon has come the realization that the "on" switches of human appetite are far more sensitive than the "off" switches. The reason for this asymmetry is that throughout the past three million years, natural selection has favored individuals who eat to store fat whenever there is an opportunity to do so. Individuals who did not eat to store fat were at a disadvantage during hungry seasons and famines. Because of the relatively few opportunities for people in hunter-gatherer societies to sustain weight gains over long periods, very little selection has occurred against those whose eating to store fat results in healththreatening obesity. This defect in the human genome may eventually be overcome by scientific and political-economic interventions aimed at controlling weight gain, but in the meantime the weak biological shut-off system is a standing invitation to the food industry to manipulate aversions and preferences regardless of health-damaging consequences.

References Cited
Abrams, L. 1979 1980 Baksh, M. 1982 The Relevance of Paleolithic Diet in Detennining Contemporary Nutritional Needs. Journal ofApplied Nutrition 31:43-59. Vegetarianism: An Anthropological/Nutritional Evaluation. Journal of Applied Nutrition 32: 53-87.

The Impact of Increased Fish and Game Scarcity on Machiguenga Subsistence Behavior. Paper read at the annual meeting of the American Anthropological Association, Washington, D. C. Bayless, T. M., and N. Rosensweig 1966 A Racial Difference in the Incidence of Lactase Deficiency. Journal of the American Medical Association 197:968-72. Binford, L. 1983 In Pursuit of the Past. New York: Thames and Hudson. Carroll, M. n.d. Why We Don't Eat Dogs-Usually. Manuscript.


I. Theoretical Overview

Chang, K. C. 1977

Introduction. In Food in Chinese Culture: Anthropological and Historical Perspectives, K. C. Chang, ed., pp. 3-21. New Haven: Yale University

Press. Cohen, M. N. 1977 The Food Crisis in Prehistory. New Haven: Yale University Press. Cuatrecasas, P.; D. H. Lockwood; and]. Caldwell 1965 Lactase Deficiency in the Adult. Lancet 1: 14-18. Darby, W.; P. Ghalioungui; and L. Grivetti 1977 Food: The Gift of Osiris. 2 vols. New York: Academic Press. DeHavenon, A. L. 1977 Rank Ordered Behavior in Four Urban Families: A Comparative VideoAnalysis of Patterns of Superordination in Two Black and White Families. Ph.D. dissertation, Columbia University. Denton, D. 1983 The Hunger for Salt: An Anthropological, Physiological and Medical Analysis. New York: Springer-Verlag. Deshler, W. W. 1965 Native Cattle Keeping in Eastern Africa. In Man, Culture, and Animals, Anthony Leeds and A. P. Vayda, eds., pp. 153-68. Washington, D. C.: American Association for the Advancement of Science. DeWalt, K., and G. Pelto 1977 Food Use and Household Ecology in a Modernizing Mexican Community. In Nutrition and Anthropology in Action, T. Fitzgerald, ed., pp. 7493. Amsterdam: Van Gorcum. Douglas, M. 1966 Purity and Danger: An Analysis ofConcepts ofPollution and Taboo. New York: Praeger. 1972 Deciphering a Meal. Daedalus 101:61-82. 1984 Standard Social Usages of Food. In Food and the Social Order: Studies of Food and Festivities in Three American Communities, Mary Douglas, ed., pp. 1-39. New York: Russell Sage Foundation. Ferro-Luzzi, G. E. 1977 Remarks on D. P. McDonald's "Food Taboos." Anthropos 73:593-94. 1980a Food Avoidances at Puberty and Menstruation in Tamiland. In Food,

Ecology, and Culture: Readings in the Anthropology ofDietary Practices,
1980b ]. Robson, ed., pp. 93-100. New York: Gordon and Breach. Food Avoidance in Pregnant Women in Tamiland. In Food, Ecology, and Culture: Readings in the Anthropology of Dietary Practices, ]. Robson, ed., pp. 101-8. New York: Gordon and Breach. Food Avoidances During the Puerperium and Lactation in Tamiland. In


Food, Ecology, and Culture: Readings in the Anthropology of Dietary Practices, ]. Robson, ed., pp. 109-17. New York: Gordon and Breach.
Flannery, K. 1968 Archaeological Systems Theory and Early MesoAmerica. In An-


2. Foodways

thropological Archaeology in the Arnericas, B. Meggars, ed., pp. 67-87. Washington, D. C.: Anthropological Society of Washington.
Garine, I. de 1978 Population, Production, and Culture in the Plains Societies of Northern Cameroon and Chad: The Anthropologist in Development Projects. Current Anthropology 19:42-65. Culture et Nutrition. Communications 31:70-90. Approaches to the Study of Food and Prestige in Savannah Tribes: Massa and Mussey of Northern Cameroon and Chad. Social Science Information 19: 39-78. Limiting Factors in Amazonian Ecology. Paper read at the annual meeting of the American Anthropological Association, Washington, D.C. Life and Death at Gambi. National Geographic 155:592-620.

1979 1980

Good, K. 1982 Goodall, ]. 1979 Goody, ]. 1982 Gross, D. 1982

Cooking, Cuisine, and Class: A Study of Comparative Sociology. New
York: Cambridge University Press. Village Movement in Relation to Resources in Amazonia. In Adaptive Responses of Native Amazonians, R. Hames and W. Vickers, eds., pp.

429-49. New York: Academic Press. Gross, D., and B. Underwood 1971 Technological Change and Caloric Costs: Sisal Agriculture in Northeastern Brazil. A merican Anthropologist 73: 725-40. Hames, R., and W. Vickers 1982 Optimal Diet Breadth Theory as a Model to Explain Variability in Amazonian Hunting. American Ethnologist 9:358-78. Hamilton, W. ]., and C. D. Busse 1978 Primate Carnivory and Its Significance to Human Diets. BioScience 28:761-66. Harris, M. 1966 The Cultural Ecology of India's Sacred Cattle. Current Anthropology 7:51-66. 1968 The Rise ofAnthropological Theo1)~ New York: T. Y. Crowell. 1977 Cannibals and Kings: The Origins of Cultures. New York: Random House. Cannibals and Kings: An Exchange. New York Review ofBooks, June 25, 1978 pp. 51-52. Cultural Materialism: The Struggle for a Science of Culture. New York: 1979 Random House. Animal Capture and Yanomamo Warfare: Retrospect and New Evidence. 1984 Journal ofAnthropological Research 40: 183-20l. 1985 Desperate Choices: Gender Hierarchies and the Diet of Pregnant and


I. Theoretical Overview Lactating Women. Paper read at the annual meeting of the American Anthropological Association, Washington, D. C. Good to Eat: Riddles of Food and Culture. New York: Simon and Schuster.

1986 Harrison, G.

Sociocultural Correlates of Food Utilization and Waste in a Sample of Urban Households. Ph.D. dissertation, University of Arizona. Hasegawa, T., et al. 1983 New Evidence on Scavenging Behavior in Wild Chimpanzees. Current Anthropology 24:231-32. Hawkes, K.; K. Hill; and j. F. O'Connell 1982 Why Hunters Gather: Optimal Foraging and the Ache of Eastern Paraguay. American Ethnologist 9:379-98. Hayden, B. 1981 Subsistence and Ecological Adaptations of Modem Hunter-Gatherers. In Omnivorous Primates: Gathering and Hunting in Human Evolution, R. S. O. Harding and G. Teleki, eds., pp. 334-422. New York: Columbia University Press. Hornabrook, R. 1977 Human Ecology and Biomedical Research: A Critical Review of the International Biological Programme in New Guinea. In Subsistence and Survival: Rural Ecology in the Pacific, T. P. Bayliss-Smith and R. G. Feacham, eds., pp. 23-61. New York: Academic Press. Hsu, V. Y. H., and F. L. K. Hsu 1977 Modem China: North. In Food in Chinese Culture. K. C. Chang ed., pp. 295-331, New Haven: Yale University Press. Hudson, C. (ed.) The Black Drink: A Native American Tea. Athens: University of 1979 Georgia. Katz, S. Une Exemple d'Evolution Bioculturelle: La Feve. Communication 1979


Food, Behavior and Biocultural Evolution. In The Psychology of Human Food Selection, L. M. Barker, ed., pp. 171-88. Westport, Conn.: AVI. Kensinger, K. M., and W. B. Kracke (eds.) Food Taboos in Lowland South America. Working Papers on South 1981 American Indians, Bennington, Vt.: Bennington College. Klausen, A. M. 1968 Kerala Fishermen. Oslo: Scandinavian University Books. Kolata, G. Food Affects Human Behavior. Science 218: 1209-10. 1983 Lee, R. The !Kung San: Men, Women, and Work in a Foraging Society. New 1979 York: Cambridge University Press. Levi-Strauss, C. 1963 Totemism. Boston: Beacon Press.



2. Foodways

1965 Le Triangle Colinaire. L'Arc 26: 19-29. Lindenbaum, S. 1977 The "Last Course": Nutrition and Anthropology in Asia. In Nutrition and Anthropology inAction, T. Fitzgerald, ed., pp. 141-55. Amsterdam: Van Gorcum. Kuru Sorcery: Disease and Danger in the New Guinea Highlands. Palo 1979 Alto: Mayfield. Lowie, R. Subsistence. In General A nthropology, Franz Boas, ed., pp. 282-326. 1938 New York: D. C. Heath. The Transitions of Civilizations in Primitive Society. American Journal of 1942 Sociology 47: 527-43. Lowie's Selected Papers in Anthroj)ology, C. DuBois, ed. Berkeley: Uni1960 versity of California Press. Culture and Ethnology. New York: Basic Books [orig. 1917]. 1966 Luzbetak, L. The Socio-Religious Significance of a New Guinea Pig Festival. An1954 thropological Quarterly 2: 59-80, 102-28. McArthur, M. 1977 Nutritional Research in Melanesia:: A Second Look at the Tsembaga. In Subsistence and Survival: Rural Ecology in the Pacific, T. P. BaylissSmith and R. G. Feacham, eds., pp. 91-128. New York: Academic Press. McDonald, D. P. 1977 Food Taboos: A Primitive Environmental Protection Agency. Anthropos 72:734-38. Maher, V. 1981 Work, Consumption, and Authority Within the Household. In Of Mar-

riage and the Market: Women's Subordination in International Perspective, K. Young et aI., eds., pp. 69--87. London: CSE Books.
Meggitt, M. 1965 The Association Between Austrahan Aborigines and Dingoes. In Man, Culture, and Animals, A. Leeds and A. Vayda, eds., pp. 7-26. Wash-

ington, D. C.: American Association for the Advancement of Science. Miller, B. 1981 The Endangered Sex. Ithaca: CornelllJniversity Press. Mintz, S. 1979 Time, Sugar and Sweetness. Marxist Perspectives 9: 56-73. Montellano, O. de 1978 Aztec Cannibalism: An Ecological Necessity? Science 200:611-17. Morren, G. 1977 From Hunting to Herding: Pigs and the Control of Energy in Montane New Guinea. In Subsistence and Survival: Rural Ecology in the Pacific, T. P. Bayliss-Smith and R. G. Feacham, eds., pp. 273-315. New York: Academic Press.


1. Theoretical Overview

Paolisso, M., and R. Sackett 1982 Hunting Productivity Among the Yukpa Indians of Venezuela. Paper read at the annual meeting of the American Anthropological Association, Washington, D. C. Rajalakshmi, S., and C. V. Ramakrishnan 1980 The Nutritional Scene in]ambudweepam (Asia). Paper read at the meeting on the Anthropology of Food and Nutrition, Bad Homburg, October 13-15, under the auspices of Maison des Sciences de I'Homme and Werner-Reimers-Stiftung. Rappaport, R. 1967 Pigs for the Ancestors. New Haven: Yale University Press. 1984 Pigs for the Ancestors. 2d ed. New Haven: Yale University Press. Ross, E. 1978 Food Taboos, Diet, and Hunting Strategy: The Adaptation to Animals in Amazon Cultural Ecology. Current Anthropology 19: 1-36. Patterns of Diet and Forces of Production: An Economic and Ecological 1980 History of the Ascendancy of Beef in the United States Diet. In Beyond the Myths of Culture: Essays in Cultural Materialism, Eric Ross, ed., pp. 181-225. New York: Academic Press. "The Political Ecology of the Irish Famine." Manuscript. 1983 Rozin, P. Human Food Selection: The Interaction of Biology, Culture, and Indi1982 vidual Experience. In The Psychobiology ofHuman Food Selection, L. M. Barker, ed., pp. 225-54. Westport, Conn.: AVI. Rozin, P., and D. Schiller 1980 The Nature and Acquisition of a Preference for Chili Pepper by Humans. Motivation and Emotion 4: 77-101. Sahlins, M. Culture and Practical Reason. Chicago: University of Chicago Press. 1976 Schneider, H. The Subsistence Cattle Among the Pakot and in East Africa. American 1957 Anthropologist 59: 278-300. Scrimshaw, N. 1981 Significance of the Interactions of Nutrition and Infection in Children. In Textbook ofPediatric Nutrition, Robert Suskind, ed., pp. 229-40. New York: Raven Press. Scrimshaw, S. Infant Mortality and Behavior in the Regulation of Family Size. In Fertil1981 ity Decline in the Less Developed Countries, Nick Eberstadt, ed., pp. 195-318. New York: Praeger. Infanticide as Deliberate Fertility Regulation. In Determinants ofFertility 1983

in Developing Countries: Fertility Regulation and Institutional Influences, R. Lee and R. Bulatoo, eds., pp. 245-66. Washington, D.C.:
National Academy of Science Press. Simoons, F. 1982 Geography and Genetics as Factors in the Psychobiology of Human


2. Foodways

Food Selection. In The Psychobiology of Human Food Selection, L. M. Barker, ed., pp. 205-24. Westport, Conn.: AVI. Soler, ].

Sopher, D. E.

The Semiotics of Food in the Bible. In Food and Drink in History, R. Foster and O. Ranum, eds., pp. 126-38. Baltimore: Penguin. Indian Civilization and the Tropical Savanna Environment. In Human Ecology in Savanna Environments, D. Harris, ed., pp. 185-207. New York: Academic Press.




Chi'Ing. In Food in Chinese Culture: Anthropological and Historical Perspectives, K. C. Chang, ed., pp. 261-94. New Haven: Yale University Press.

Speth, ].

Sponsel, L.

Bison Kills and Bone Counts. Chicago: University of Chicago Press.
Yanomama Warfare, Protein Capture, and Cultural Ecology: A Critical Analysis of the Arguments of the Opponents. Interciencia 8: 204-10. Oral and Facial Innate Motor Response to Gustatory and Some Olfactory Stimuli. In Preference Behavior and Chemoreception, ]. Kroeze, ed., pp. 247-62. London: Information Retrieval.

Steiner, ].


Teleki, G. The Omnivorous Diet and Eclectic Feeding Habits of Chimpanzees in Gombe National Park, Tanzania. In Omnivorous Primates: Gathering and Hunting in Human Evolution, R. S. O. Harding and G. Teleki, eds., pp. 303-43. New York: Columbia University Press. Thompson, A., and F. Hytten 1977 Physiological Basis of Nutritional Needs During Pregnancy and Lactation. In Nutritional Impacts on Women Throughout Life, K. Moghisi and T. Evans, eds., pp. 10-22. New York: Harper and Row. Turner, M. R. (ed.) 1981 Preventive Nutrition and Society :New York: Academic Press. Vaidyanathan, A.; K. N. Nair; and M. Harris 1982 Bovine Sex and Species Ratios in India. Current Anthropology 23:365-


Vayda, A.; A. Leeds; and D. Smith 1961 The Place of Pigs in Melanesian Subsistence. In Proceedings of the 1961 Annual Spring Meeting of the American Ethnological Society, pp. 69-77. Seattle: University of Washington Press. Wagley, C. 1977 Welcome of Tears. New York: Oxford University Press. Welsch, R. 1981 An Interdependence of Foodways and Architecture: A Foodways Contrast on the American Plains. In Food in Perspective: Proceedings of the

Third International Conference on Ethnological Food Research, Cardiff,


1. Theoretical Overview

Wales, 1977, Alexander Fenton and Trefor Owen, eds., pp. 263-376. Edinburgh: John Donald.
Wilson, C.


Food Taboos of Childbirth: The Malay Example. In Food, Ecology, and Culture: Readings in the Anthropology of Dietary Practices, ]. Robson, ed., pp. 67-74. New York: Gordon and Breach.

Winterhalder, B., and E. A. Smith (eds.)

Wolf, E.

Ifunter-Gatherer Foraging Strategies: Ethnographic and Archaeological Analyses. Chicago: University of Chicago Press. Europe and the People Without History. Berkeley: University of California Press.

Wood, C.


Human Sickness and Health: A Biocultural View. Palo Alto: Mayfield.


Bioevolutionary Antecedents and

LIKE MOST PRIMATES, HUMANS AR1~ OMNIVORES IN THE NARrow sense of consuming both plants and animals. This omnivorous capability is undoubtedly a phylogenetic characteristic-a ,part of "human nature." Yet omnivory is a rather vague dietary concept. Non-human primates display considerable species-specific variation in the ratio ofplants to animals in the diet. Primate foodways also differ in the ratio ofparticular kinds ofplants or plant parts (e.g., roots, leaves, fruits, seeds) and in the kinds of animal materials consumed (e.g., insects, vertebrates). The question arises, therefore, as to whether human beings have a phylogenetic"legacy" that leads us to pre.fer animal overplant foods (or vice versa) or to prefer certain kinds ofplants over others and certain kinds of animal materials over others. The essays that follow address this question. The logical first place to look for an answer is in the human gut. To the extent that human foodways are a phylogenetic legacy, one would expect them to be associated with phylogenetically distinct morphological and physiological features. As Milton shows in Chapter 3, human gut morlJhology and physiology are poorly designed for processing large quantities ofbulky, fibrous plant material. This does not necessarily mean that the human digestive tract has been selected for an ability to digest large amounts of animal flesh. Indeed, Hamilton's essay (Chapter 4) explicitly denies this possibility, citing the dangers posed by the overconsumption of meat in contemporary societies. Hatnilton's data on chacama baboons suggest instead that primate preferences for animal foods (mostly insects) are opportunistic and can largely be accounted for by the more favorable energy input/out ratio associated with meat (a theme that is explored in greater detail in Part IV). While Milton and Hamilton are concerned ,with pan-human phylogenetic constraints on human foodways, the possibility that intra-specific genetic polymorphisms account for variations in foodways associated with particular popula-


II. Bioevolutionary Antecedents and Constraints

tions must also be kept in mind. This possibility is addressed in Katz's study ofthe role offava beans in circum-Mediterranean foodways (Chapter 5). Like the betterknown case of selection for adult lactase sufficiency and milk consumption as a buffer against calcium-deficiency diseases, the ability to consume fava beans also appears to have been selected for the biochemical protection the beans provide against malaria. Although Katz's gene-culture co-evolution model possesses considerable cogency with respect to a number of important food preferences and aversions, and is especially welcome as the explanation of apparently irrational foodways that cannot be explained culturologically, it seems unlikely that the vast majority of human intra-specific food preferences and avoidances can be traced to such one-to-one culture-gene associations.



Primate Diets and Gut Morphology: Implications for Hominjd Evolution
THERE IS CURRENTLY STRONG INTER~EST IN DEVELOPING A BETter understanding of the probable food habits and dietary niche of early humans (Isaac 1978; Peters and O'Brian 1981; Stahl 1984). Without such infonnation, we are handicapped in our ability to interpret the significance of many features of human morphology and to construct viable models of early human ecology (Isaac 1978; Sussman 1978). Further, it is increasingly obvious that many of the major health problems faced today by more modem technological societies stem from factors related to diet. This strongly suggests that the average diet in such societies is not entirely suitable for human nutritional needs (Burkitt, Walker, and Painter, 1972; Trowell 1978; Truswell 1977). In this paper, I review infonnation about dietary choice in primates, paying particular attention to members of the Hominoidea. I then examine features of the human gut, comparing it with the guts of other mammals, both primates and non-primates, to distinguish any features that appear to set humans apart. I conclude by speculating on the probable diet of early humans, using the behavior of extant pongids as a partial foundation for my speculations.

Humans are generally viewed as omnivores (Fischler 1981; Harding 1981). By definition, an omnivore is any animal that takes food from more than one trophic level. Most mammals are in fact omnivorous (Landry 1970; Morris and Rogers 1983a, 1983b), including such diverse forms as pigs, tayras, dogs, pandas, bears, primates, skunks, some bats, and dozens of rodents. Though all omnivorous mammals appear to have basically similar nutritional requirements, different species satisfy these needs in different ways, using a tremendous


II. Bioevolutionary Antecedents and Constraints

range and variety of foods. Describing a given species as omnivorous, therefore, does little to clarify what foods it may depend on, or how its digestive capabilities may differ from those of other omnivores, or why it may show decided preferences for some food types over others. Both pure herbivores, such as bovids, and pure carnivores, such as felids, show highly characteristic metabolic specializations to the peculiar end products of their respective and specialized diets (Morris and Rogers 1983b). In bovids, microbial fermentation results in little glucose being absorbed from the gut. Ruminant metabolism is adapted toward the use of acetate rather than glucose for fatty acid synthesis and a constant high rate of gluconeogenesis (Morris and Rogers 1982; Van Soest 1982). Further, because bovids typically obtain all essential amino acids and most vitamins from gut flora, they do not have to choose foods to meet these needs. Similarly, pure carnivores, with strict adherence to a diet of animal tissue, absorb little glucose as a result of digestive processes and show a pattern of carbohydrate metabolism differing from that of omnivores (Morris and Rogers 1982). Pure carnivores have lost the ability to synthesize certain proteins (enzymes) that appear to be of no advantage to them because of their highly specialized diets. Adult cats, for example, require a dietary source of argenine, an amino acid typically synthesized by adult omnivores in sufficient amounts to meet maintenance requirements (Morris and Rogers 1982, 1983a, 1983b). In general, the pure carnivore appears intolerant of diets adequate for the adult omnivore (ibid.). Humans are not ruminants; nor do they possess the suite of specialized metabolic adaptations to diet that distinguishes the pure carnivore. Thus, in terms of metabolic adaptations to diet, humans fall in with other omnivorous mammals. To clarify more precisely what type of omnivory may have characterized ancestral hominids, it is useful to examine certain broad characteristics of the primate diet, paying special attention to hominoids.

The Primate Diet
The adaptive radiation and eventual dominance of angiosperms during the Cretaceous opened up a variety of new dietary opportunities (Regal 1977). Potential foods included not only insects that pollinated angiosperm flowers but also the pollen, nectar, fruits, seeds, and foliage of the angiospenns themselves. The primate line is believed to have differentiated by the Middle Paleocene, arising from some type of terrestrial insectivorous stock (Eisenberg 1981). If present-day primates are any indication, early primates appear to have taken strong advantage of arboreal plant foods. All extant primates take food



Primate Diets and Gut Morphology

from the first trophic level, but not all primates take food from the second, at least not intentionally. A very few prosimians (e.g., Galago demidoviz) take the bulk (wet weight) of their diet from animal nlatter (Charles-Dominique 1977), but the overwhelming majority of primates take the bulk (wet weight) of the diet from plants, eating only small amounts of animal matter (Gaulin and Konner 1977; Harding 1981; Hladik 1977). This indicates that the adaptive radiation of primates, particularly the anthropoids, occurred by virtue of their ability to penetrate the as yet unfilled arboreal plant food niche and radiate to the point where they came to dominate a strong subset of the available arboreal dietary resources. In the tropical forest, almost all potential plant food comes from dicotyledonous species using the C3 carbon path'way. In sharp contrast, many potential plant foods in the savannas come from monocotyledonous species using the C4 carbon pathway. I do not know whether these differences in plant food types are reflected to any degree in the digestive physiology of primates. Certainly, if extant primates are any indication, the primate gut was initially adapted for both the nutritive and the defensive components of dicotyledonous C3 rather than monocotyledonous C4 plant foods. Recent experimental work on human fiber digestion shows that human microflora are very sensitive to different fiber sources. As a group, humans are very efficient at degrading the relatively unlignified hemicelluloses and cellulose of dicot vegetable fibers such as cabbage or carrots but are less efficient on monocot cereal fibers such as wheat bran or monocot plant fibers such as alfalfa, with a high cellulose to hemicellulose ratio and considerable lignification (Van Soest et al. 1983). Though most extant primates eat primarily dicotyledonous plant species, a few species (e.g., Papio spp., Theropithecus gelada) eat quantities of grass blades and presumably are able to degrade the dietary fiber of grasses with some degree of efficiency. The current dependence of most large human populations on quantities of monocotyledonous plant foods, particularly cereal grains, is a notable departure from the traditional plant foods consumed by the majority of primates, both in the past and today. Monocot cereal grains also tend to be high in phytate, which, because of its high anionic character, is ideal for forming complexes with mineral elements, particularly transitional elements such as zinc, iron, and manganese (Lloyd, McDonald, and Crompton 1978). Any primate turning to phytate-rich plant foods as a major dietary staple may require other special foods in the diet to avoid potential mineral deficiencies due to complexing of phytates with essential minerals. Animal foods, for example, are a good source of zinc, which could help to augment losses due to binding of this element in cereal foods by phytates. In choosing foods, small primates, because of the increase in the ratio of metabolic requirements to gut capacity (Demment and Van Soest 1985), tend


II. Bioevolutionary Antecedents and Constraints

to select rapidly digestible plant foods with little bulk, such as sugary fruit or gum (Gaulin and Kanner 1977; Hladik 1977). These high-quality plant foods are supplemented to a greater or lesser extent by animal matter, typically insects. Larger-bodied primates generally include some foliage in the diet, eating leaves and buds as well as fruits, seeds, gum, flowers, and cambium. When eating foliage, most primates focus their attention on young rather than mature leaves in spite of the fact that mature leaves are far more abundant. Younger leaves generally show a higher protein-to-fiber ratio than mature leaves and tend to be less lignified, thus offering greater nutritive returns to the feeder (Milton 1979). Some larger-bodied primates routinely seek out foods from the second trophic level, particularly insects, but occasionally vertebrates (e.g., Cebus spp., Papio spp., Pan troglodytes), whereas others rarely appear to ingest animal matter intentionally (e. g., A louatta spp.). A wealth of available data indicates that primates, particularly anthropoids, typically include a number of different plant parts and plant species in the daily diet. Over an annual cycle, for example, howler monkeys (Alouatta palliata) living in lowland tropical forest on Barra Colorado Island, Republic of Panama, take foods from more than 109 plant species (Milton 1980). Further, on any given day, howlers may take foods from 10 to 20 or more different plant species. Two other primate species sympatric with howlers in this same forest, the black-handed spider monkey (Ateles geoffroyz) and the capuchin monkey (Cebus capucinus), also take foods from well over a hundred plant species per annum, using 10 or more plant species per day (Hladik and Hladik 1969; Oppenheimer 1968; Milton, unpublished data). Arboreal and semi-arboreal animals of some other orders do not appear to include as many plant species in the diet either per annum or per day. For example, the Barra Colorado forest has also been the site of detailed study of the dietary ecology of the red-tailed tree squirrel (Sciurus granatensis), two-toed sloth (Choloepus hoffmanni), threetoed sloth (Bradypus variegatus), and coati (Nasua narica) (Glanz et al. 1982; Montgomery and Sunquist 1978; Russell 1979). These species range from purely herbivorous (the two sloth species) to omnivorous (squirrel and coati); in no case do individuals of any of these species even begin to approach the dietary diversity reported for the monkey species in this forest. The large number of plant species eaten by most primate species is an interesting and perhaps unique characteristic of the primate diet (but see Sussman 1978 for a contrasting view). Conversely, scant data suggest that primate species routinely including animal prey in the diet tend to show strong prey specificity within particular habitats (see, e.g., the work of Charles-Dominique [1977] on the Gabon lorisids). The broad plant food niche of primates and the apparently more narrow and specialized animal food niche are an aspect of primate food choice that warrants further study and quantification.


3. Primate Diets and Gut Morphology

The Hominoid Diet
An examination of the food choices of hylobatids and pongids shows that all species conform to the general primate trend just described in that they eat considerable plant material, supplementing it with some animal matter, typically eaten in small or even trace amounts. Mountain gorillas are almost exclusively herbivorous (Fossey and Harcourt 1977; Goodall 1977), followed by orangutans (Rodman 1977, 1984) and siamangs (Chivers 1977; Gittins and Raemaekers 1980), both of which eat notable amounts of leaves, shoots, stems, and/or bark as well as fruit and some insect matter (Table 3.1). Gibbons and chimpanzees focus very strongly on fruit in the diet, eating some foliage and from 4 percent to 13 percent animal matter (Gittins and Raemakers 1980; Rodman 1984; Wrangham 1977). The diet of the pygmy chimpanzee is not as yet well documented, but it appears to feed partially on fruit and partially on fibrous vegetable matter, particularly that of ground cover species, supplemented by some animal matter (Kano 1983). As a dietary category, ripe fruit tends to be relatively high in soluble carbohydrates but low in protein, whereas leaves tend to be relatively high in protein but low in soluble carbohydrates (Milton 1979, 1981). Primate species routinely including large quantities of foliage in the diet each day would not be expected to take foods from the second trophic level unless such foods were required for some essential trace nutrient such as vitamin B 12 .

The Primate Digestive Tract
The above data show that primates are omnivores of a particular type in that the great majority show a clear focus on plant foods, eating only modest amounts of animal matter. Generally, however, when we think of plant-eating animals, primates do not come to mind; rather, we think of cows, sheep, horses, koalas, kangaroos, and the like. All of these highly herbivorous forms routinely eat the phytosynthetic tissues of plants as their staple item of diet, and all show digestive tracts that are amazingly specialized in form when compared with those of most primates. As an order, primates show a digestive tract that, in its general form, is not greatly rnodified from the primitive mammalian pattern (Mitchell 1905). In many respects the anthropoid gut is simpler in form than that of prosimians, and it has been suggested that their respective gut forms may have been independently derived from the primitive mammalian pattern (Mitchell 1905). The relatively unspecialized form of the normative primate gut supports the view that, as an order, primates, particularly the


II. Bioevolutionary Antecedents and Constraints

TABLE 3.1. Food Choices of Hominoids (Percentage of Feeding Time)a
Leaves, Shoots, and Stems

Gorilla gorilla beringei
(mountain gorilla)


Other Plant Foods
2 flowers 7 wood 3 roots

Animal Matter


Pongo pygmaeus

53 (often unripe) 63 (ripe) 68

25 20 28

15 bark

<1 insects 3 insects

b b

Pan troglodytes
(common chimpanzee)


4 insects, eggs, fledglings b termites(? )

Pan paniscus
(pygmy chimpanzee)


Hylobates syndactylus



6 flowers


Hylobates lar
Oar gibbon)

50 58

29 39 Plant foods, ?60

7 flowers 3 flowers


Hylobates agilis
(agile gibbon)

Homo sapiens
(modem hunter-gatherers' excluding highest latitudes)

ca. 30-40


SOURCES: a, Fossey and Harcourt (1977); b, Rodman (1984); c, Hladik (1977); d, Kano (1983); e, Gittins and Raemaekers (1980); f, Lee (1968). aValues should not necessarily sum to 100 percent as different sources may not present data on all food types eaten by their study subjects. bChimpanzees were also observed to hunt mammals at Gombe site (Teleki 1981).

anthropoids, have traditionally focused on very high-quality plant foods that are not extensively fibrous or lignified and supplemented them with some secondtrophic-level foods. Primates deviating from this pattern of gut morphology appear to represent special radiations that have turned heavily to foliage as a dietary staple (i. e., Colobinae, Indriidae).


.3. Primate Diets and Gut Morphology

It should be noted that within lineages the ancestral pattern tends to dominate gut fonn (Mitchell 1905). Gut fonn, therefore, cannot in itself be used to predict diet. For example, a well-developed cecum is generally associated with diets high in plant fiber. All extant prosimians have a cecum, in most cases capacious, but have not been noted to eat any great amount of plant fiber. In many cases the prosimian cecum appears to function as a fennentation chamber for plant exudates and/or chitin of insect exoskeletons (Clemans 1980; Sheine 1979). Conversely, if one examines animals from other orders, the hippopotamus and the giant panda are both strongly herbivorous, and both lack a cecum (Hill and Rewell 1948; P. j. Van Soest:, pers. comm.). The bear, a decided omnivore, has a gut as simple in fonn as that of the mink, a strong carnivore (Hill and Rewell 1948). A sacculated stomach and the strong development of both the cecum and colon are generally good predictors of diets high in plant fiber. But the absence of such features does not mean that a given species is

TABLE 3.2. Relative Gut Volume Proportions for Some Hominoid Species (Percentage of Total Volume)
Source of Raw Data
a a a a a b

Gorilla gorilla

25 17 20 24 24 17 24 10

Small Intestine
14 28 23 25 29 67 56 63

7 3 5

53 54 52 49

Pongo pygmaeus

Pan troglodytes

Hylobates syndactylus

Hylobates pileatus
(pillated gibbon)


45 17 19 23

Homo sapiens

Homo sapiens Homo sapiens

1 3


SOURCES: a, Chivers and Hladik (1979); b, Maynard and Loosli (969); c, D. J. Chivers, pers. comm.; d, R. W. Sussman, pers. comm. All calculations of relative volumes by the author. NOTE: These figures are not scaled with respect to inter-specific differences in body size and as such should only be used inter-specifically as an indication of the pattern of gut proportions of the different species. As data are often taken from immature specimens or single specimens, all of these data, perhaps excluding those for Homo sapiens, should be regarded as rough estimates.


II. Bioevolutionary Antecedents and Constraints

not strongly herbivorous, nor does it mean that a given species is strongly carnivorous. Staple items of diet or even trophic levels may alter within members of a particular family or genus, but, very broadly, phylogeny appears to dominate function in terms of gut form (Mitchell 1905). This is not to imply that all members of a particular lineage share the same gut scale or proportions, for such is most emphatically not the case. Within lineages, many features of the gut can show modification, particularly in the volume or length of particular sections (see Table 3.2). Work on a number of bird and mammal species likewise shows that within species, between individuals, and perhaps even in the same individual, the nature of the diet can affect gut scale and proportions (Gentle and Savory 1975; Gross, Wang, and Wunder, in press; Koong et al. 1982; Miller 1975; Moss 1972; Murray, Tulloch, and Winter 1977). There are also presumed to be a number of more subtle differences within lineages and intra-specifically in terms of morphological and physiological features of the gut such as mucosa thickness, villi length, and the like (Hill 1949; Hladik 1967; Karasov and Diamond 1983; Sonntag 1924).

The Hominoid Digestive Tract
In keeping with the above observation, all members of the Hominoidea show the same basic gut pattern. Hominoids have a simple acid stomach, a small cecum terminating in a true appendix, and a well-sacculated colon. The hominoid appendix represents a shared trait of this superfamily not found in other extant primates. The night monkey (Aotus spp.) is reported to show some development of an appendix, though not to the same extent as hominoids (Hill and Rewell1948). The hominoid appendix appears to represent the culmination of a strong trend in primates for lymphoid tissue to collect in the cecal apex (Berry 1900; Hill and Rewell 1948). In hominoids, this lymphoid tissue has migrated from the cecal apex into a discrete structure. The functions of the hominoid appendix are not known, but it is clearly an active, functioning organ (Hill and Rewell1948). In humans, the appendix secretes an alkaline fluid containing amylase, eripsin, and mucin. Davenport (1971) suggests that the human appendix is an especially lively site of antibody production. In rabbits, the appendix has been found to serve immunological functions, producing antibodies against certain protein antigens (Draper and Sussdorf 1965; Hanaoka, Nomoto, and Waksman 1970; Konda and Harris 1966). Other species with an appendix are the capybara (Hydrochoerus hydrochaeris) and the wombat (Lasiorhinus latifrons). The rabbit, capybara, and wombat, like all hominoids, are strongly herbivorous. Though the basic form of the hominoid gut is similar throughout the super-


3. Primate Diets and Gut Morphology

family, there are notable differences between humans and other hominoids when relative gut proportions are compared (Table 3.2). Humans concentrate by far the greatest gut volume in the small intestine (2::56 percent), whereas gibbons and orangutans show the greatest gut volume in the colon. In addition, the size of the human gut relative to body mass is small in comparison with most other anthropoids (R. D. Martin, pers. comm.). A variety of animal studies indicate that increases in energy requirenlents without a decrease in dietary quality will increase the size of the small intestine and decrease the colon (Cripps and Williams 1975; Fell, Smith, and 'Campbell1963; Gross, Wang, and Wunder, in press). Certainly the present-day gut proportions of humans in modem technological societies indicate utilization of nutritionally dense, energetically concentrated foods. At the present time, however, it is difficult to evaluate the implications of differences in gut proportions between humans and other hominoids. As noted, many animal species are able to respond rapidly to changes in dietary quality in terms of modification of gut proportions. The size of the present-day human small intestine could be an ancient or a relatively recent trait. Indeed, it is not known whether all modem human populations show such gut proportions. On average, for example, individuals in Western societies are estimated to take in no more than 10 grams of fiber per day, whereas members of some rural African populations may take in more than 170. A difference of this magnitude, in view of the gut plasticity demonstrated for animals of some other orders, suggests that there may be some differences in gut proportions between extant human populations. Of interest would be data on the gut proportions of human populations on different dietary substrates as well as data on the plasticity of the hominoid gut in terms of its responses to changes in diet. Burkitt, Walker, and Painter (1972), in observing defecation patterns of rural Ugandans eating unrefined, fibrous diets, cornmented that they showed transit times two to five times as rapid as those of British navy personnel eating refined Western diets and produced some four to five times more fecal matter. A number of other studies of human fiber digestion likewise show that increasing the fiber level of the diet (which generally implies lowered dietary quality) significantly decreases mean transit time (e. g., Wrick et al. 1983). A similar response was found in chimpanzees fed trial diets of different fiber levels (Milton and Demment, in prep.). Extrapolating to the natural environment, this pattern of digestive kinetics indicates that \vhen dietary quality in the natural environment declines, both humans and chin1ps respond by increasing intake, which results in a more rapid turnover rate of ingesta. This kinetic response could help to ensure that individuals of both species continue to supply energy and nutrients to body tissue at an optimal rate in spite of some fluctuations in the availability of higher-quality foods in the natural environment. The relatively small size of the human colon appears to represent the de-


II. Bioevolutionary Antecedents and Constraints

rived rather than ancestral condition of Hominoidea. The colon of human neonates is more similar to that of pongids than is the case for mature individuals (Hill 1949). Humans show regression of the colon as they mature, whereas pongids show elongation, particularly of the left colon. Hill (1949) points out that in the arrangement of the colon, as in many other features of anatomy, pongids appear to have a gerontomorphic status as compared with humans. The marked sacculation of the human colon can also be viewed as a possible retention feature. When compared with those of most other mammals, the relative proportions of the human gut are unusual (my calculations, using data from Chivers and Hladik 1980 and Hladik 1967). Pure carnivores, such as felids, or more omnivorous carnivora, such as canids, do not have gut proportions similar to humans'. Rather, more carnivorous animals tend to show considerable volume in the stomach or in the stomach and small intestine. For example, 70 percent of the gut volume of the adult cat occurs in the stomach, whereas for dogs this figure is 62 percent. HigWy herbivorous forms also differ from humans. Ruminants tend to show the greatest volume in the region of the stomach, whereas non-ruminant herbivores such as equines show tremendous volume in the cecum and colon. Swine, often regarded as good omnivore analogues for humans, in fact differ considerably. Some swine have a specialized area in the stomach near the pyloric region that is totally lacking in humans; further, swine are characterized by a large cecum relative to that of humans and a tremendous proportion of gut volume in the lower tract (EWe et al. 1982; Stanogias and Pearce 1985a, 1985b). Work by EWe et al. (1982) suggests that pigs may also have a somewhat different pattern of lower-gut turnover than humans as a result of cecal pulsing. When the relative proportions of the human gut are compared with those of other primates, it is still difficult to find a good match. Most anthropoids show notable volume in the cecum and/or colon or have a higWy specialized stomach (e.g., Colobinae). Prosimians show gut proportions somewhat similar to those of humans in that the small intestine tends to dominate the gut. However, like some carnivora, some prosimians show greater relative volume in the stomach than is the case for humans; further, most prosimians have a notable cecum that in some cases is highly specialized (Clemens 1980). One primate whose gut proportions are strikingly similar to those of humans is the New World capuchin monkey (Cebus spp.). Like humans, capuchins concentrate most gut volume in the small intestine. The ratio of gut mass to body mass in capuchins is also small in comparison with other non-human anthropoids (R. D. Martin, pers. comm.). Capuchin monkeys eat a high-quality diet made up of unusually rich wild foods, both sugary fruits and protein- and oil-rich seeds. Capuchins also routinely devote 40 to 50 percent of their daily foraging time to seeking out second-trophic-Ievel foods, including soft-bodied


3. Primate Diets and Gut Morphology

grubs, cicadas, and small vertebrates (Oppenheimer 1968; Parker and Gibson 1977; Terborgh 1983; Milton, unpublished data). Scant data on one specimen of Papio papio suggest that this species may be somewhat similar to humans in relative gut proportions (Chivers and Hladik 1980; Hladik 1969). Here too there appears to be an emphasis on the volume of the small intestine relative to other sections of the gut. Savanna baboons, like capuchin monkeys, are unusually selective feeders who specialize in high-quality foods. Baboons at times may devote alrnost all of their daylight hours to painstakingly seeking out small, nutritious food items such as corms, acacia gum, grass seeds, flowers, fruits, and animal matter, including copious quantities of insects when these are available (DeVore and Hall 1965; Hamilton, Buskirk, and Buskirk 1978). Baboons also feed opportunistically on small vertebrates, particularly immature animals (De\lore and Hall 1965; Harding 1973; Strum 1981). Capuchin monkeys and savanna baboons are also unusual primates in that both use the hand to a considerable degree both to find and to prepare food items for consumption (Beck 1975; Parker and Gibson 1977). Capuchins are noted for their manual dexterity. They routinely use the hand to crack hardshelled fruits, to unroll dead leaves in search of insect prey, and to pry among palm fronds for insects and small vertebrates (Parker and Gibson 1977). Baboons rely heavily on the hand when feeding, particularly to remove dirt from food items, to peel, husk, and open food items, and to grasp live prey. M. W. Demment (pers. comm.) points out that an adult male baboon (Papio cynocephalus) may pick up as many as 3,000 individual food items in a single day of feeding, each weighing no more than one-tenth of a gram dry weight. Thus, savanna baboons are heavily committed to the use of the hand in feeding. The frequent modification of a dietary item before ingestion in effect buffers the teeth and digestive tract of these species from the physical effects of many items in their diet. The similarity in relative gut proportions of humans, capuchin monkeys, and perhaps savanna baboons is not derived from a close common ancestor. Rather, it appears to represent similar adaptive trends in gut morphology in response to diets made up of unusually high-quality dietary items that are capable of being digested and absorbed primarily in the small intestine.

Food Choices of Early Humans
As discussed above, the comparative anatolny of the hominoid digestive tract indicates that modern human gut proportions and scale represent the derived rather than the ancestral condition for the superfamily Hominoidea. There can


II. Bioevolutionary Antecedents and Constraints

be little doubt that the ancestral line giving rise to this superfamily and ultimately to hominids was markedly herbivorous. Kliks (1978) has presented evidence from analyses of human coprolites to document the fact that until quite recently, many human populations took in an impressive amount of plant fiber in the diet, estimated from rehydration at perhaps some 130 grams of fiber per day. In addition, human coprolites also contain undigested residues of animal tissue, including such materials as bones, teeth, hair, feathers, keratinized skin, fish scales, and insect cuticle that at times contribute more than 10 percent of the total weight of undigested residues (Kliks 1978). Without more data on the comparative gut proportions of modem human populations and the degree of short-term gut plasticity characteristic of humans and apes, it is difficult to state whether the gut proportions and scale of modem humans as reported in this paper in fact characterized early humans. However, my prediction is that all extant humans will be found to have a gastrointestinal tract dominated by the small intestine, though considerable variation may be recorded for the size of the colon region. Examination of the diets and activity patterns of extant pongids, in combination with evidence from the hominid fossil record, suggests that early humans focused feeding on energy-rich, high-quality foods. With an adult body weight of 93 kilograms (female) to 160 kilograms (male), gorillas are by far the largest anthropoids (Clutton-Brock and Harvey 1977). Gorillas typically feed on quantities of leafy material, a dietary category that is low in soluble carbohydrates. On such a diet, large body size confers decided energetic and nutritional advantages. In mammalian herbivores and omnivores, relative gut capacity and body mass show a linear relationship, whereas maintenance metabolism and body mass show an exponential relationship (Demment and Van Soest 1985; Parra 1978). In effect, a larger herbivore has proportionately more room in the gut and can exploit foods with a lower protein/fiber ratio than its smaller-sized counterparts (Demment and Van Soest 1985; Janis 1976; Nuzum 1985; Parra 1978). The large body size of the gorilla facilitates exploitation of a fibrous dietary matrix, but on such a diet energy appears to be in short supply. Gorillas are relatively inactive for terrestrial anthropoids and also show low levels of social interaction. This suggests that energy may be limited, so that behaviors that conserve energy are favored. Similarly, orangutans, though often described as fruit-eaters, in fact concentrate much of their feeding on unripe fruit, leaves, and bark (Rodman 1977). Like gorillas, orangutans tend toward a fibrous dietary substrate that is often presumed to be low in soluble carbohydrates, and, again like gorillas, they are relatively immobile. Ninety percent of their travel takes place at a slow pace through the trees; the average day range is only some 300 meters (Rodman 1977). Further, orangutans are relatively unsocial and are one of the few anthropoid species not associating in any type of relatively permanent social group.


3. Primate Diets and Gut Morphology

In contrast to gorillas and orangutans, chimpanzees focus their feeding primarily on high-quality foods. The keystone of the chimpanzee diet is ripe fruit, and individuals are very active, often traveling three to four kilometers a day in search of sufficient ripe fruit to meet nutritional requirements. Over 70 percent of chimpanzee travel takes place on the ground. Chimps use smaller, more dispersed food sources than orangutans and show longer median interpatch distances between food sources in ~10-minute feeding bouts (Rodman 1984). They supplement their basic ripe fruit diet \vith young leaves of unusually high quality (Hladik 1977) as well as insect and mammal prey. Male chimpanzees have been observed to hunt monkeys and pJlgS, and meat resulting from communal kills may be shared. (However, it should be noted that no more than 6 percent of the total annual diet of chimpanzees is estimated to come from second-trophic-Ievel foods; see Table 3.1.) Chirnpanzees are also extremely social and have a rich repertoire of facial gestures and calls. When feeding, chimpanzees make use of stones to crack hard-shelled fruits (Boesch and Boesch 1981) and use twigs and grasses to harvest termites (McGrew 1974). Thus, like capuchin monkeys and savanna baboons, chimpanzees rely on the hand for many fine-level manipulations with respect to food preparation, and in their case tools may also be employed (see e.g., Parker and Gibson 1977). In summary, these comparative data on the dietary foci and behaviors of extant pongids strongly suggest that in the hominoid line, a focus on lowerquality, more fibrous plant foods leads to selection for a larger-bodied and relatively inactive and unsocial primate. In pongids, there is a clear pattern toward increasing the relative size of the hindgut and increasing the fiber content of the diet with increasing body size (Milton and Demment, in prep.). Early humans are believed to have evolved in a savanna-mosaic setting. High-quality foods, both plant and animal, are more patchily distributed in both space and time in a savanna environment than in tropical forests (Milton 1981; Milton and May 1976; D. Olson, pers. comm.). This implies that early hominids in such a setting may have had both large day ranges and large home ranges if they concentrated on higher-quality, more digestible foods. In the hominid line, bipedalism is a more energetically efficient terrestrial locomotor mode than quadrupedalism (Rodman and McHenry 1980). Rodman and McHenry (1980) have hypothesized that selective pressures related to increased travel efficiency between widely dispersed food sources in a savanna setting may underlie in the adaptation of bipedalism in the hominid line. One way to lower foraging costs when moving from the tropical forest into a savanna-mosaic setting is simply to lower dietary quality. The "robust" australopithecines may have opted for this type of dietary strategy. Their relatively large post-canine teeth and massive skull bones suggest that they fed on tough, fibrous, and/or hard plant foods (G~rine 1981). M. Demment (pers. comm.) suggests that dietary competition between the digestively specialized ungulates and the robust australopithecines may have contributed to the


II. Bioevolutionary Antecedents and Constraints

eventual extinction of the robust forms. However, another way to lower foraging costs when moving into a patchy savanna environment is to continue to specialize in high-quality foods and to cover increased foraging costs both by improving food-search efficiency and by eating even higher-quality foods (Milton 1980, 1981). This may have been the dietary strategy of the "gracile" australopithecines. Ultimately, however, an adaptive peak should be reached, such that no further improvement on this basic foraging strategy can occur, because there are a finite number of ways to locate a finite number of highquality dietary items efficiently in the savanna. All australopithecines, both robust and gracile, are characterized by thick molar enamel and large cheek teeth. Australopithecines show somewhat more cranial expansion than extant pongids, but the difference is slight (Holloway 1973; Leutenegger 1973). In contrast, members of the genus Homo show thinner molar enamel, a dramatic reduction in cheek tooth size, and considerable cranial expansion (Grine 1981; McHenry 1982; S. Ambrose, pers. comm.). In combination, these dental and cranial features, as well as an increase in body size, apparently with no loss of mobility or sociality, strongly imply that early members of the genus Homo made a dramatic breakthrough with respect to diet-a breakthrough that enabled them to circumvent the nutritional constraints imposed on body size increases in the apes. It would appear that early humans were able in some manner to greatly improve their intake and uptake of energy, apparently without any decrease in dietary quality. Such a dietary breakthrough had to go beyond improved food search efficiency or simple utilization of available high-quality foods, for I think that the gracile australopithecines were probably already at the apex of possibilities in this respect. There had to be some type of novel innovation-either technological or social or both-that altered the dietary potential of proto-humans (see, e. g., Lancaster 1968, 1975). A technological innovation could somehow make a lowquality but available and abundant food into a high-quality food. A social innovation, such as cooperative hunting and food sharing, could make formerly inaccessible or restricted high-quality food accessible and relatively dependable. An innovation such as language could help to coordinate foraging activities and thereby greatly enhance foraging efficiency (see, e. g., Lancaster 1968, 1975). I cannot state what this innovation was, but perhaps it was both technological and social, for certainly we see the strong development of both trends in human evolution. I can, however, speculate on possible selective pressures. For most of its evolutionary history, the ancestral line leading to hominoids presumably lived in a forested environment. Plant foods are presumed to have composed the bulk of the diet, complemented perhaps by a modest amount of second-trophic-Ievel food (Kliks 1978; Milton 1981, 1984). Data from extant hunter-gatherer societies suggest that this basic pattern of primate omnivory may also have been practiced by most hunter-gatherer groups living in tropical


3. Primate Diets and Gut Morphology

areas (Lee 1968). It is only in temperate to .Arctic latitudes that second-trophiclevel foods are noted at times to compose the bulk of the diet (Lee 1968). Early hominids and humans are believed to have evolved in the tropics. If this assumption is correct, it is doubtful, by analogy with both the diets of extant primates and what is known of the diets of extant tropical hunter-gatherer societies, that animal protein in itself composed the bulk of the early human diet (Hayden 1981; Speth and Spielmann 1983). Indeed, research suggests that for most modem humans large quantities of animal protein may actually be detrimental to both normal growth and good health (Edozien and Switzer 1978; Nelson 1975). Human populations such as the Arctic Eskimo, whose diet is composed primarily of animal matter, show special adaptations for energy and nitrogen metabolism, and it is speculated that some of their dietary adaptations may be under some degree of genetic control (Draper 1977). Further, Arctic Eskimos do not eat a diet of pure animal protein but rather eat a mixture of animal protein and animal fat. Animal protein seems most appropriate in the human diet when it is eaten in combination "rith notable amounts of either fat or carbohydrates and used primarily to meet demands for amino acids and nitrogen (Edozien and Switzer 1978; Maynard and Loosli 1969; Nelson 1975). It seems unlikely that animal protein has ever served as the principal item of diet for the majority of tropical-living human populations. Given the patchy nature of higher-quality foods in a savanna environment, however, I would suggest that both animal protein and animal fat may have been important dietary resources for early humans. Though relatively few nonhuman primate species live in the savannas, those that do frequently include animal matter in the diet, at times in considerable amounts (Hamilton, Buskirk, and Buskirk 1978). This suggests that in a savanna environment, animal matter may be somewhat more available for larger-bodied primates than it is in a tropical forest. If early humans were able to depend on protein-rich animal foods to fulfill their daily amino acid requirements, this would buffer the digestive tract from selective pressures related to the need to efficiently process large quantities of proteinacious plant matter-typically leaves. A larger body mass could perhaps be supported with less gut mass, as is suggested to be the case for some carnivores as well as capuchin monkeys and modem humans (Chivers and Hladik 1979; R. D. Martin, pers. com~m.). Routinely using some animal matter in the diet would make proportionately more room available in the gut to process carbohydrate-rich plant foods, the traditional energy source for the great majority of primate species. Plant foods differ in many important respects from animal foods, placing the plant-eating animal under somewhat different selective pressures than carnivores with respect to features of foraging success (Milton 1984; Westoby 1974). Plant foods are sessile and tend to be buffered from consumption by internal, chemical characteristics such as low nutrient content, high cell wall


II. Bioevolutionary Antecedents and Constraints

matter, or secondary compounds. In contrast, animal prey is typically highly mobile and protected from predation by external defenses such as speed, spines, teeth, or claws (Milton 1984). If early humans devoted some foraging effort each day to the procurement of second- as well as first-trophic-Ievel foods, their foraging strategy, necessarily focused for most of their past evolutionary history on the efficient exploitation of sessile plant foods, must have undergone some rather radical modifications. The pressures to become efficient at procuring foods from two rather than one trophic level may have set in motion a new suite of behaviors (see, e. g., Strum 1981), leading eventually to what we recognize as the Homo grade of development. Like some other researchers (e. g., Isaac 1978; Lancaster 1968, 1975), I see a division of labor with respect to food procurement in combination with food sharing as a pivotal adaptation in human evolution. Indeed, I think that the implications of this type of dietary innovation have not been fully appreciated, for, in effect, a division of labor and food sharing provide a means whereby individuals of a given species can efficiently utilize foods from two trophic levels simultaneously-a foraging strategy that appears to be truly unique among mammals.

Portions of this paper were presented at the 1984 Gordon Conference on Evolution of the Human Diet in the session titled "Comparative Studies of Modem Feeding Systems." As the paper that I presented at the Wenner-Gren-sponsored Food Preferences and Aversions Conference was already committed to another volume, I am pleased to be able to publish this material in its stead. I thank David Chivers and Robert Sussman for generously sharing with me their original data on the gut proportions of modem humans. Conversations with Peter Van Soest and Montague Demment contributed greatly to many ideas in this paper.

References Cited
Beck, B. 1975 Baboons, Chimpanzees and Tools. Journal of Human Evolution 3:50916.

Berry, R. ]. A. 1900 The True Caecal Apex or the Vermiform Appendix: Its Minute and Comparative Anatomy. Journal ofAnatomy and Physiology 35:83-100. Boesch, C., and H. Boesch 1981 Sex Differences in the Use of Natural Hammers by Wild Chimpanzees: A Preliminary Report. Journal ofHuman Evolution 10:585-93.


3. Primate Diets and Gut Morphology

Burkitt, D. P.; A. R. P. Walker; and N. S. Painter 1972 Effect of Dietary Fiber on Stools and Transit Times and Its Role in the Causation of Disease. Lancet 2: 1408-1I. Charles-Dominique, P. 1977 Ecology and Behavior ofNocturnal Primates. New York: Columbia University Press. Chivers, D. j. 1977 The Feeding Behaviour of Siamang (Symphalangus syndactylus). In Primate Ecology, T. H. Clutton-Brock, ed., pp. 355-413. London: Academic Press. Chivers, D. j., and C. M. Hladik 1980 Morphology of the Gastrointestlllal Tract in Primates: Comparisons with Other Mammals in Relation to Diet. Journal of Morphology 166:337-86. Clemens, E. rr. 1980 The Digestive Tract: Insectivore, Prosimian and Advanced Primate. In Comparative Physiology: Primitive Mammals, K. Schmidt-Nielson, L. Bolis, and C. R. Taylor, eds., pp. 90-99. Cambridge: Cambridge University Press. Clutton-Brock, T. H., and P. H. Harvey 1977 Species Differences in Feeding and Ranging Behaviour in Primates. In Primate Ecology, T. H. Clutton- Brock, ed., pp. 557-84. Cambridge: Cambridge University Press. Cripps, A. W., and V.]. Williams 1975 The Effects of Pregnancy and Lactation on Food Intake, Gastrointestinal Anatomy and the Absorptive Capacity of the Small Intestine in the Albino Rat. British Journal o/Nutrition 33: 17-32. Davenport, H. W. 1971 Physiology of the Digestive Tract. 3d ed. Chicago: Year Book Medical Publishers. Demment, M. W., and P. ]. Van Soest 1985 A Nutritional Explanation for Body-Size Patterns of Ruminant and NonRuminant Herbivores. A merican Naturalist 125: 641- 72. DeVore, I., and K. R. L. Hall 1965 Baboon Ecology. In Primate Behavior, I. DeVore, ed., pp. 20-52. New York: Holt, Rinehart and Winston. Draper, H. H. 1977 The Aboriginal Eskimo Diet in Modern Perspective. American Anthropologist 79: 309-16. Draper, L. R., and D. H. Sussdorf 1965 Roles of the Liver and Appendix in the Serum Hemolysin in Rabbits. Journal ofImmunology 95:306-13. Edozien, j. C., and B. R. Switzer 1978 Influence of Diet on Growth in the Rat. Journal of Nutrition 108:28290.


II. Bioevolutionary Antecedents and Constraints

EWe, F. R.; ]. L. Jeraci; ]. B. Robertson; and P.]. Van Soest 1982 The Influence of Dietary Fiber on Digestibility, Rate of Passage and Gastrointestinal Fermentation in Pigs. Journal of Animal Science 55: 1071-80. Eisenberg, ]. F. 1981 The Mammalian Radiations. Chicago: University of Chicago Press. Fell, B. F.; K. A. Smith; and R. M. Campbell 1963 Hypertrophic and Hyperplastic Changes in the Alimentary Canal of the Lactating Rat. Journal ofPathology and Bacteriology 85: 179-88. Fischler, C. 1981 Food Preferences, Nutritional Wisdom and Sociocultural Evolution. In Food, Nutrition and Evolution, D. N. Walcher and N. Kretchmer, eds., pp. 59-68. New York: Masson. Fossey, D., and A. H. Harcourt 1977 Feeding Ecology of Free-Ranging Mountain Gorilla (Gorilla gorilla beringet). In Primate Ecology, T. H. Clutton-Brock, ed., pp. 415-47. London: Academic Press. Gaulin, S. ]. C., and M. Konner 1977 On the Natural Diet of Primates, Including Humans. InNutrition and the Brain, vol. 1, R. ]. Wurtman and]. ]. Wurtman, eds., pp. 1-86. Ne"w York: Raven Press. Gentle, M. ]., and C. ]. Savory 1975 The Effects of Dietary Dilution on the Intestinal Anatomy of the Japanese Quail (Coturnix coturnixjaponica). Research in Veterinary Science 19:284-87. Gittins, S. P., and]. ]. Raemaekers 1980 Siamang, Lar and Agile Gibbon. In Malayan Forest Primates, D. ]. Chivers, ed., pp. 63-106. New York: Plenum Press. Glanz, W. E., et al. 1982 Seasonal Food Use and Demographic Trends in Sciurus granatensis. In The Ecology of a Tropical Forest, E. G. Leigh, A. S. Rand, and D. M. Windsor, eds., pp. 239-52. Washington, D. C.: Smithsonian Press. Goodall, A. G. 1977 Feeding Behaviour of a Mountain Gorilla Group (Gorilla gorilla beringet) in the Tshibinda-Kahuzi Region (Zaire). In Primate Ecology, T. H. Clutton-Brock, ed., pp. 450-79. London: Academic Press. Grine, F. E. 1981 Trophic Differences Between "Gracile" and "Robust" Australopithecines: A Scanning Electron Microscope Analysis of Occlusal Events. South African Journal ofScience 77: 203-30. Gross, J.; Z. Wang; and B. A. Wunder In press Adaptations to Food Quality and Energy Needs: Changes in Gut Morphology and Capacity of Microtus ochrogaster. Journal ofMammalogy. Hamilton, W. ].; R. E. Buskirk; and W. H. Buskirk 1978 Omnivory and Utilization of Food Resources by Chacma Baboons, Papio

ursinus. American Naturalist 112:911-24.


3. Primate Diets and Gut Morphology

Hanaoka, M.; K. Nomoto; and B. H. Waksman 1980 Appendix and M-Antibody Formation I: Immune Response and Tolerance to Bovine Globulin in Irradiated, Appendix-Shielded Rabbits. Journal ofImmunology 104:616-25. Harding, R. S. O. 1973 Predation by a Troop of Olive Baboons (Papio anubis). American Journal ofPhysical Anthropology 38:587-92. 1981 An Order of Omnivores: Nonhuman Primate Diets in the Wild. In Omnivorous Primates: Gathering and Hunting in Human Evolution, R. S. O. Harding and G. Teleki, eds., pp. 191-214. New York: Columbia University Press. Hayden, B. 1981 Subsistence and Ecological Adaptations of Modem Hunter/Gatherers. In Omnivorous Primates: Gathering and Hunting in Human Evolution, R. O. Harding and G. 1'eleki, eds., pp. 344-421. New York: Columbia University Press. Hill, W. C. O. 1949 Some Points in the Enteric Anatomy of the Great Apes. Proceedings of the Zoological Society ofLondon 119: 19-32. Hill, W. C. 0., and R. E. Rewell 1948 The Caecum of Primates: Its Appendages, Mesenteries and Blood Supply. Transactions of the Zoological Society ofLondon 26: 199-257. Hladik, A., and C. M. Hladik 1969 Rapports Trophiques Entre Vegetation et Primates dans la Foret de Barro Colorado (Panama). Terre et Vie 23:25-117. Hladik, C. M. 1967 Surface Relative du Tractus Digestif de Quelques. Primates: Morphologie des Villosites Intestinales et Correlations Avec Ie Regime Alimentaire. Mammalia 31: 120-47. Adaptive Strategies of Primates in Relation to Leaf-Eating. In The Ecol1977 ogy of Arboreal Folivores, G. G. Montgomery, ed., pp. 373-96. Washington, D. C.: Smithsonian Press. Holloway, R. L. 1973 Endocranial Volumes of Early African Hominids and the Role of the Brain in Human Mosaic Evolution. Human Evolution 2:449-59. Isaac, G. 1978 Food Sharing and Human Evolution: Archaeological Evidence from the Plio-Pleistocene of East Africa. Journal of Anthropological Research 34:311-25. Janis, C. 1976 The Evolutionary Strategy of the Equidae and the Origins of Rumen and Cecal Digestion. Evolution 30: 757-74. Kano, T. 1983 An Ecological Study of the Pygmy Chimpanzee (Pan paniscus) of Yalosidi, Republic of Zaire. International Journal of Primatology 4: 132.


II. Bioevolutionary Antecedents and Constraints

Karasov, W. H., and]. M. Diamond 1983 Adaptive Regulation of Sugar and Amino Acid Transport by Vertebrate Intestine. American Journal of Physiology 245 (Gastrointestinal Liver Physiology 8): G443-62. Kliks, M. 1978 Paleodietetics: A Review of the Role of Dietary Fiber in Preagricultural Human Diets. In Topics in Dietary Fiber Research, G. A. Spiller and R. ]. Amen, eds., pp. 181-202. New York: Plenum Press. Konda, S., and T. N. Harris 1966 Effect of Appendectomy and of Thymectomy, with X-irradiation, on the Production of Antibodies in Two Protein Antigens in Young Rabbits. Journal ofImmunology 97:805-14. Koong, L. ]., et a1. 1982 Effects of Plane of Nutrition on Organ Size and Fasting Heat Production in Pigs. Journal ofNutrition 113: 1626-31. Lancaster, ]. B. 1968 Primate Communication Systems and the Emergence of Human Language. In Primates: Studies in Adaptation and Variability, P. C. Jay, ed., pp. 439-57. New York: Holt, Rinehart and Winston. 1975 Primate Behavior and the Emergence of Human Culture. New York: Holt, Rinehart and Winston. Landry, S. O. The Rodentia as Omnivores. Quarterly Review ofBiology 45:351-72. 1970 Lee, R. B. 1968 What Hunters Do for a Living; or, How to Make Out on Scarce Resources. In Man the Hunter, R. B. Lee and 1. DeVore, eds., pp. 30-48. Chicago: Aldine Press. Leutenegger, W. 1973 Encephalization in Australopithecines: A New Estimate. Folia Primatolica 19: 9-17. Lloyd, L. E.; B. E. McDonald; and E. W. Crompton 1978 Fundamentals ofNutrition. 2d ed. San Francisco: W. H. Freeman. McGrew, W. C. 1974 Tool Use in Wild Chimpanzees in Feeding Upon Driver Ants. Journal of Human Evolution 3:501-8. McHenry, H. M. 1982 The Pattern of Human Evolution: Studies on Bipedalism, Mastication and Encephalization. Annual Review ofAnthropology 11: 151-73. Maynard, L. A., and]. K. Loosli 1969 Animal Nutrition. New York: McGraw-Hill. Miller, M. R. Gut Morphology of Mallards in Relation to Dietary Quality. Journal of 1975 Wildlife Management 39: 168-73. Milton, K. Factors Influencing Leaf-Choice by Howler Monkeys: A Test of Some 1979 Hypotheses of Food Selection by Generalist Herbivores. American N aturalist 114: 362-78.


3. Primate Diets and Gut Morphology


The Foraging Strategy ofHowler Monkeys: A Study in Primate Economics. New York: Columbia University Press.

Distribution Patterns of Tropical Plant Foods as an Evolutionary Stimulus to Mental Development in Primates. American Anthropologist 83:534-48. 1984 The Role of Food Processing Factors in Primate Food Choice. In Adaptations for Foraging in Nonhuman Primates, P. Rodman and J. Cant, eds., pp. 249-79. New York: Columbia University Press. Milton, K., and M. W. Demment In prep. Digestive Kinetics and Assimilation Efficiencies of Chimpanzees (Pan troglodytes): A Model for Human Fiber Digestion. Milton, K., and M. L. May 1976 Body Weight, Home Range and Diet in Primates. Nature 259:45962. Mitchell, P. C. 1905 On the Intestinal Tract of Mammals. Transactions of the Zoological Society ofLondon 17:437-536. Montgomery, G. G., and M. E. Sunquist 1978 Habitat Selection and Use by Two-Toed and Three-Toed Sloths. In The Ecology of Arboreal Folivores, G. G. Montgomery, ed. pp. 329-60. Washington, D.C.: Smithsonian Press. Morris, J. G., and Q. Rogers 1982 Metabolic Basis for Some of the Nutritional Peculiarities of the Cat. Journal ofSmall Animal Practice 23: 599-613. 1983a Nutritionally Related Metabolic Adaptations of Carnivores and Ruminants. In Plant, Animal and Microbial Adaptations to Terrestrial Environment, N. S. Margaris, M. Arianoutsou-Faraggitaki, and R. J. Reiter, eds., pp. 165-80. New York: Plenum. 1983b Nutritional Implications of Some Metabolic Anomalies of the Cat. In


American Animal Hospital Association 50th Annual Meeting Proceedings, pp. 325-31. San Antonio: AAHA.
Moss, R. 1972 Effects of Captivity on Gut Lengths in Red Grouse. Journal of Wildhfe

Management 36: 99-104.
Murray, D. M.; N. M. Tulloch; and W. H. Winter The Effect of Three Different Growth Rates on Some Offal Components 1977 of Cattle. Journal ofAgricultural Science (Cambridge) 89: 119-28. Nelson, R. A. 1975 Implications of Excessive Protein. In Proceedings Western Hemisphere Nutrition Congress IV, P. L. White and N. Selvy, eds., pp. 71-76. Acton, Mass.: Publishing Sciences Group. Nuzum, C. T. 1985 Morphological Correlations. Science 229: 428. Oppenheimer, ]. R. 1968 Behavior and Ecology of the White-Faced Monkey (Cebus capucinus) on Barro Colorado Island. Ph. D. dissertation, University of Illinois, Urbana.


II. Bioevolutionary Antecedents and Constraints

Parker, S. T., and K. R. Gibson 1977 Object Manipulation, Tool Use and Sensorimotor Intelligence as Feeding Adaptations in Cebus Monkeys and Great Apes. Journal ofHuman Evolution 6: 623-41. Parra, R. 1978 Comparison of Foregut and Hindgut Fermentation in Herbivores. In The Ecology of Arboreal Folivores, G. G. Montomgery, ed., pp. 205-30. Washington, D. C.: Smithsonian Press. Peters, C. R., and E. M. O'Brian The Early Hominid Plant Food Niche: Insights from an Analysis of Plant 1981 Exploitation by Homo, Pan and Papio in Eastern and Southern Africa. Current Anthropology 22: 127-40. Regal, P. ]. 1977 Ecology and Evolution of Flowering Plant Dominance. Science 196:62229. Rodman, P. S. 1977 Feeding Behaviour of Orang-utans of the Kutai Nature Reserve, East Kalimantan. In Primate Ecology, T. H. Clutton-Brock, ed., pp. 384-414. London: Academic Press. 1984 Foraging and Social Systems of Orang-utans and Chimpanzees. In Adaptations for Foraging in Nonhuman Primates, P. S. Rodman and]. Cant, eds., pp. 134-60. New York: Columbia University Press. Rodman, Peter S., and H. M. McHenry 1980 Bioenergetics and the Origins of Hominid Bipedalism. AmericanJournal ofPhysical Anthropology 52: 103-6. Russell, ]. K. 1979 Reciprocity in the Social Behavior of Coatis (Nasua narica). Ph.D. dissertation, University of North Carolina, Chapel Hill. Sheine, W. S. 1979 Digestibility of Cellulose in Prosimian Primates. American Journal of Physical Anthropology 50:480-81. Sonntag, C. F. 1924 The Morphology and Evolution of the Apes and Man. London: John Bale Sons and Danielsson. Speth, ]. D., and K. A. Spielmann 1983 Energy Source, Protein Metabolism, and Hunter-Gatherer Subsistence Strategies. Journal ofAnthropological Archaeology 2: 1-31. Stahl, A. 1984 Hominid Diet Before Fire. CurrentAnthroplogy 25:151-68. Stanogias, G., and G. R. Pearce 1985a The Digestion of Fiber by Pigs 1: The Effects of Amount and Type of Fibre on Apparent Digestibility, Nitrogen Balance and Rate of Passage. British Journal ofNutrition 53:513-30. 1985b The Digestion of Fibre by Pigs 2: Volatile Fatty Acid Concentrations in Large Intestine Digesta. British Journal ofNutrition 53: 531-36. Strum, S. 1981 Processes and Products of Change: Baboon Predatory Behavior at


3. Primate Diets and Gut Morphology

Gilgil, Kenya. In Omnivorous Primates: Gathering and Hunting in Human Evolution, R. S. O. Harding and G. Teleki, eds., pp. 255-302. New York: Columbia University Press. Sussman, R. W. 1978 Foraging Patterns of Nonhuman Primates and the Nature of Food Preferences in Man. Federation Proceedings, Anthropology and the Assessment ofNutritional Status 37:55-60. Teleki, G. The Omnivorous Diet and Eclectic Feeding Habits of Chimpanzees in 1981 Gombe National Park, Tanzania. In Omnivorous Primates: Gathering and Hunting in Hunlan Evolution, R. S. O. Harding and G. Teleki, eds., pp. 303-43. New York: Columbia University Press. Terborgh, ]. Five New World Primates: A Study in Comparative Ecology. Princeton: 1983 Princeton University Press. Trowell, H. 1978 The Development of the Concept of Dietary Fiber in Human Nutrition. A merican Journal of C finical Nutrition 31: S3-S11. Truswell, A. S. 1977 Diet and Nutrition of Hunter-Gatherers. In Health and Disease in Tribal Societies, pp. 213-26. Aba Foundation Symposium no. 49 (new series). New York: Elsevier. Van Soest, P. ]. 1982 Nutritional Ecology of the Ruminant. Corvallis, Oreg.: 0 & B Books. Van Soest, P. ]., et al. 1982 Comparative Fermentation of Fibre in Man and Other Animals. Paper read at the International Symposium on Dietary Fiber, Palmerston, North, New Zealand. 'Van Soest, P. ]., et a1. 1983 Some in vitro and in vivo Properties of Dietary Fiber from Noncereal Sources. ACS Symposium Series, no. 214, Unconventional Sources of Dietary Fibre, I. Furda, ed., pp. 135-141. Washington, D. C.: American Chemical Society. Westoby, M. 1974 An Analysis of Diet Selection by Large Generalist Herbivores. American Naturalist 108: 290-304. Wrick, K. L.; ]. B. Robertson; P. ]. Van Soest; B. A. Lewis; j. M. Rivers; D. A. Roe; and L. R. Hackler 1983 The Influence of Dietary Fiber Source on Human Intestinal Transit and Stool Output. Journal ofNutrition 113: 1464-79. Wrangham, R. W. 1977 Feeding Behaviour of Chimpanzees in Gombe National Park, Tanzania. In Primate Ecology, T. H. Clutton-Brock, ed., pp. 504-38. London: Academic Press.



Omnivorous Primate Diets and Human Overconsumption of Meat

preferences is an aggregate science, consisting of several distinct approaches. These include: (1) analysis of early hominid diets; (2) the study of the diets of non-human primate species and possible regulating principles underlying them; and (3) theories concerned with patterns of foraging by animal species in general. What is the relationship of our dietary past to present human dietary predilections? Here I briefly evaluate some evidence concerning early hominid diets and the diets of certain contemporary ground-dwelling primates. The dietary habits of contemporary primates may help qualify our interpretation of evidence about past and present hominid diets.

Legacy Hypotheses
Some explanations of contemporary human behavior are based upon early hominid or pre-hominid traits, thus implying adaptive inertia-that is, biological evolution less rapid than environmental change. Such hypotheses suggest that some traits are not adaptive under current conditions and are legacies from earlier evolution. Inertia or legacy explanations for behavior may be more relevant to humans than to other animal species because of the rapidity with which humans have changed their own environments. (But changes in patterns of human behavior may also be more rapid than biological evolution if cultural change is involved, a possibility considered later in this paper.) Inertia explanations have been offered to explain various human characteristics. For example,


II. Bioevolutionary Antecedents and Constraints

Orians (1980) and Balling and Falk (1982) suggest that the contemporary habitat and overall environment of human choice, expressed today by lawn and garden styles, resembles the early African savanna landscape inhabited by our prehistoric ancestors. These authors speak of a generalizable human preference based upon an adaptive response to that environment. Conceding that humans like what they are familiar with and what is advertised or preferred by neighbors, they contend nevertheless that there is an innate preference for a short-cropped, grassy landscape, and that this preference emerged in prehistoric African savannas. The proponents of such legacy arguments do not reject the influence of evolutionary and/or cultural responses to changed environmental and social conditions. Rather, they insist that some part of the evolutionary past, near or remote, remains incorporated into the human behavioral repertoire. EiblEibesfeldt (1979) and many others (e.g., Alexander 1971; Bigelow 1969) use the same sort of argument when they suggest that humans once lived in small bands, and that patterns of social support, including the expression of group hostility, originated in such groups. These patterns, they conclude, linger today in our behavioral repertoire even though such behavior may sometimes be disadvantageous to contemporary individuals and to entire populations. I am using a similar argument when I suggest that overconsumption of meat and sugar by some contemporary human populations may be based in part upon the dietary predilections of our omnivorous ancestors. Since some contemporary individuals and entire human populations eat no meat at all, a legacy argument cannot be a global explanation for current attitudes toward particular foods. Some of the differences between populations and individuals are related to cultural phenomena, a topic discussed below. Analysis of adaptive inertia, the expression in contemporary populations of traits adaptive to ancestral conditions, is not easily made. The amount of evolutionary baggage we carry may be quite different for various behavior patterns, depending upon the rigor of natural selection upon them through time and upon their malleability. Thus, without specific knowledge of the genetic and cultural bases for contemporary behavioral predispositions, it may be difficult to evaluate the extent to which a behavior pattern has wholly or in part been determined by our prehistoric past. Evolutionary scenario-spinners can suggest hypotheses far more easily than empiricists can test them. Reconstruction of behavioral prehistory and application of findings to contemporary societies cannot be as rigorous a science as one would like. In many senses it is an imperfect skill involving inferred comparisons of past and present populations. In the case of humans, the evidence comes from physical and cultural anthropology, comparative animal behavior, and, indeed, any relevant source. For past hominid populations, only indirect and fragmentary evidence remains, and it is inevitably subject to conflicting explanations and to revision as new paradigms and information emerge.


4. Omnivorous Primate Diets

Genetic Lags in Human Dietary Adaptation
Foods available to contemporary human populations are, to a large degree, the result of relatively recently developed food sources. Information about the genetic basis of human abilities to digest various foods, insofar as it exists, suggests considerable variability. Some inter-population genetic variation in human digestive capacities reflects historic differences in the availability of various kinds of foods. For example, Simoons (1981) shows that celiac or wheat-eating disease (dietary intolerance of wheat flour) occurs with increasing frequency in populations more recently exposed to wheat as a dietary staple. On a smaller geographic scale, McNicholl and co-workers (1981) make the same point. In the familiar example of lactose, high incidence of tolerance is correlated with a history of pastoral lifestyles (Kretchmer 1972; Simoons 1970). Population differences in tolerance of alcohol and other dietary novelties suggest that human capacities to utilize new foods and withstand drugs without ill effect have evolved during recent millennia. By comparison, there are abiding constraints that have not responded to cultural changes in dietary need and availability. The inability of humans and other primates to synthesize vitamin C is one example. In spite of the enormous amount of disease and death caused by vitamin C deficiency, or scurvy (Kretchmer 1981), humans, unlike many other mammalian species, remain unable to synthesize this vitamin. Humans, it would seem, have a limited ability to adjust to new dietary opportunities they have either chanced upon or created. During recent millennia and centuries, the cross-cultural spread of new foods into various human populations has resulted in diet-related health problems. What time scale are we considering when we explain contemporary dietary behavior in terms of our evolutionary legacies? The answer depends upon largely unmeasured rates of change, cultural or genetic. Does our failure to lnake optimal use of particular dietary items reflect our genetic history as tree shrews, australopithecines, or Homo erectus, or does it reflect an even more recent past? Since not all traits are subject to the same degree of selective pressure or are equally responsive to selection, any generalization about the significance of behavioral legacies needs to take into account everything we know about the sequence of previous events. For animal matter consumption-and for human diets in general-the available record leaves us with ambiguous evidence (Isaac and Crader 1981).

Cultural Change
<:ulture is an elusive process, extraordinarily difficult to analyze, since biological theories of cultural evolution are a novelty and are inadequately tested and


II. Bioevolutionary Antecedents and Constraints

refined. At one end of the spectrum is the suggestion that classical natural selection bears a close relationship to cultural processes (e. g., Durham 1976; Lumsden and Wilson 1981). It follows from this position that a search for the adaptive significance of culturally determined phenomena will be successful. If this were so, the analytical problem would be greatly simplified because we could eventually identify the relationship between contemporary nutritive practices and attitudes and inclusive fitness. But if, as is more likely the case, cultural selection follows different rules and is partially or wholly divorced from recent natural selection, the evaluation of hypotheses relating diet to past and present evolutionary adaptation will be more difficult. What I see as an untenable position is the conclusion, tacit or explicit, that culture is an entirely human phenomenon and that it can be studied without considering evolutionary influences. Here I adopt an intermediate position: although the capacity for culture has evolved, specific cultural practices have not, at least not in a classical Darwinian sense. If, as Boyd and Richerson (1982, 1983) suggest, what we have inherited is a set of cultural learning rules, then inheritance and culture may lead to practices that are not necessarily in the best interests of the individuals and populations practicing them. In this sense legacy arguments remain relevant. If the rules governing our practice and inheritance of culture evolved in an environment quite different from contemporary conditions, it is possible that cultural practices were once more closely tuned to inclusive fitness maximization. Thus, contrary to one prevailing opinion-that culture is a process allowing for rapid matching of behavior to benefit or optimization-it is also possible to conclude that culture impedes rapid evolutionary change, particularly when the environment is changing as rapidly as are contemporary human diets. Perhaps at a slower pace of environmental change, culture could more closely match practice to inclusive fitness than it has in recent millennia. This view of culture, correct or not, reconciles human culture with natural selection. The capacity for culture has evolved in those few species-notably early humans-who encountered the most variable environments. Humans are at the extreme end of the spectrum of species in terms of the amount of environmental change they are subject to. The degree to which human behavior is determined by their social milieu is exceptional when they are compared with other animal species, and, as noted above, there is no reason to assume that specific expressions of culture are adaptive. In fact, there is extensive evidence suggesting that this is not the case-for example, the observation that diet-related diseases multiplied with the development of agriculture. In what is now Illinois, pre-agricultural people showed levels of tooth wear not evident after maize agriculture emerged; however, a higher level of dental caries characterizes the later populations (Cook and Buikstra 1973). In the case of human social structure, a distinction has been made between ecologically and culturally imposed monogamy (Alexander et al. 1979). The


4. Omnivorous Primate Diets

same distinction can be applied to human diets. Food choices based upon environmental constraints may be distinguished from those which are cultural phenomena. Thus, many of the dietary peculiarities of various contemporary human populations can be considered to be culturally imposed. It follows that culture is a potential extra-evolutionary cause of maladaptive dietary practices. Nevertheless, I reject the suggestion that biological evolution is irrelevant once we conclude that a particular dietary practice is culturally determined.

Primate Omnivory and Contemporary Human Diets
'[he extrapolation from omnivorous primate to man requires accepting some untested and possibly untestable assumptions. If we assume that human ancestors possessed in general form the dietary predilections of contemporary omnivorous primates, it follows that with improved hunting and other procurelnent abilities, the proportion of animal matter in human diets could and probably did increase. The quantity of animal matter and products in the diets of some contemporary people can then be understood in terms of a nearly unlimited availability of animal matter superimposed upon food category preferences acquired in the past. According to this hypothesis, some humans are led by inherited biases to choose a diet that our ancestors sought but could not have obtained for any protracted interval. This does not mean that the search was not profitable to ancestral hominids, and I will show below how it may be advantageous to primates, which, nevertheless, incorporate animal matter into their diets at a rate that makes it, on the average, only a minor part of their diet. This is another argument for changed environmental conditions and a genetic system that has lagged in its tracking of environmental change. Since there is considerable evidence suggesting evolutionary adaptation in human digestive abilities, what is the case for the inertia argument? I would argue that we must consider each human dietary change and adaptation to it as a separate case. In some cases a simple enzyme difference determines human digestive capability. ()ther adjustments are not so simple, perhaps requiring complex changes in behavior not readily made in evolutionary time. Perhaps the most compelling evidence in favor of the inertia argument is that we suffer ill effects from some foods that nevertheless figure prominently in our diet. One such case is meat, as epidemiological evidence clearly shows. Some individuals and populations are overconsuming meat and fat to the detriment of their health, and the inertia argument offers one explanation for this biologically anomalous situation. This is a strong argument against the suggestion that nutritional wisdom (summarized by Fischler 1981) effectively guides all of the quantitative and qualitative details of human diets. However, some caution is necessary relative to this


II. Bioevolutionary Antecedents and Constraints

conclusion because the diseases associated with meat overconsumption have their greatest impact upon post-reproductive individuals. Is meat overconsumption really reducing inclusive fitness? Sugar overconsumption can be explained in the same way. Fruits are the next choice after animal matter for omnivorous primates. But ripe and sweet fruit is only seasonally available, and much fruit is eaten by baboons and other primates when it has not reached sugar maturity. This is not necessarily by choice, but baboons are not orchardists. They must compete with one another, as well as with bats, pigeons, and a host of other animals, for the fruits of their forests. Sour fruit is better than none at all. For baboons and many other omnivorous creatures, a competitive world precludes dependence upon the fruit equivalent of sugared cereals and candy. But given the occasional choice, many primates consume quantities of ripe and sweet fruit and will linger in a tree loaded with ripe fruit for prolonged intervals.

Primate Omnivory
Omnivory, as used here, refers to the propensity to consume and process three broad classes of food: leafy material, fruits and seeds, and animal matter. By this definition humans are omnivores, although not all human individuals choose this full range of food types. Omnivory represents a vast dietary difference from carnivory, frugivory, and folivory. Although lions sometimes eat grass, and folivorous and frugivorous mammals may ingest some animal matter, the bulk of their diet is largely limited to one class of foods. For the omnivores, there are fluctuations, often seasonal, in the proportions of these classes of foods in the diet. Like human diets, the diets of free-ranging primate populations differ widely. Highly folivorous primate diets are less relevant to our understanding of human diets than are those of the omnivores. Omnivorous primates choose categories of food broadly overlapping those accepted by contemporary humans. Intraand inter-specific differences in the diets of omnivorous primates may be due to (1) heritable differences, (2) conditions of seasonal availability, (3) socially induced limitations upon food availability, and (4), in a limited number of cases, culturally learned and transmitted methods for processing foods. Because of my field research experience with chacma baboons, Papio uysinus, I focus initially on the dietary characteristics of populations of this species and on the bases of individual and population differences in diet. I then discuss observations of animal matter consumption by this species, its social and ecological correlates, and the relevance of animal matter consumption by


4. Omnivorous Primate Diets

omnivorous primates to meat overconsumption by some contemporary human populations. Dietary habits of savanna baboons are considered here as one model for the ecological determinants of early human food consumption. However, it is doubtful that any contemporary human integrated into a social group could survive for long on a baboon's diet. Some of the items require longer processing times than humans might be able to afford. Others are highly distasteful or toxic. So I am not suggesting that baboons are precise primate counterparts to either contemporary or prehistoric humans. But, like some other omnivorous primates, especially the ground-d\velling macaques, langurs, and chimpanzees, they are probably the best primate models, and perhaps the best animal models, for humans and their dietary choices. Principles determining the dietary intake of savanna baboons may provide insights into the origins and consequences of omnivory in primates, including humans. Folivorous primates excepted, there is an inverse correlation between the amount of animal matter in primate diets and body size (Hamilton and Busse 1978). The basis for this correlation is also discussed by Kay and Hylander (1978) and Kay (1984). These authors conclude that it would be impossible for a large primate to satisfy its total energy requirements on an insect diet. Larger primates, even though they may prefer animal matter to other kinds of food, consume vegetation and fruits and seeds as the bulk of their diet during most seasons (Hamilton, Buskirk, and Buskirk 1978). There is a simple explanation for the unfulfilled carnivorous appetites of the larger primates. They are incapable of capturing adults of prey species as large as or larger than themselves. Moreover, when they do catch small mammals or birds, their largest prey, high-ranking individuals may monopolize access to thse items (Hamilton and Busse 1982). Occasionally insect outbreaks make animal matter plentiful and easily available. When this happens the proportion of dietary animal matter soars (Hamilton, Buskirk, and Buskirk 1978). Evidence provided in Table 4.1 shows that one large primate, the chacma baboon, weighing up to 30 kilograms or more, occasionally subsists largely upon animal matter, sometimes for protracted intervals, if this resource is sufficiently abundant and available. A similar dietary response to animal matter abundance is seen in many contemporary human populations. But in the usual day-to-day search for food, insects and larger animals are captured infrequently by baboons and other omnivorous priInates, and in most cases a large primate does not have an advantage over a smaller one in catching an individual insect. This applies both within and between species. Insects are usually seized one at a time, processed if necessary (for example, by eliminating the gut contents of large caterpillars), and then consumed. The outbreak quantified in the last column of Table 4. 1 involved a single species of caccid scale insect. These insects and their shells were


II. Bioevolutionary Antecedents and Constraints

TABLE 4.1. Chacma Baboon Food Preferences in Percentages of Time Spent Eating Foods in Each Class
No Insect Outbreak Five Troops
0.7% 77.3% 21.8%

Insect Outbreak Troop A
2.3% 84.2%

Animal matter (including insects) Fruits and seeds Vegetation Euphorics Soil

Troop H
72.1% 20.3% 7.3%:



0.20/0 0.1%



SOURCE: Hamilton, Buskirk, and Buskirk (1978). NOTES: Percentages of time five chacma baboon troops in Namibia and Botswana spent eating foods in broad classes were averaged over a one-year period. The second and third columns compare intake by two Botswana troops of the same population over a two-month interval when scale insects were abundantly available to one troop (H). The animal matter category consists largely of insects. T = trace.

scraped in groups from leaves by the baboons. They have a soft chitinous exoskeleton and are swallowed whole. Presumably their ingestion would not leave identifiable microwear striations on baboon teeth (see below). Hamilton and Busse (1978) extrapolated these and other observations of chacma baboon foraging to other omnivorous primates, including humans. The rationale for doing so is that other omnivores also attempt to maximize the rate of caloric intake, and, when readily available, animal matter is the most energyrich food. This reasoning assumes that the primary basis for choice within the range of acceptable dietary items is energy maximization. If animal matter provides the greatest net energy yield per unit foraging time, it will be the chosen category of food. Food selection by animals in general may be based upon optimization of net energy gain, within the range of acceptable items (Pulliam 1974; Pyke, Pulliam, and Charnov 1977; Schoener 1971), most nutrient needs being incidentally met in the process. Surely omnivores have dietary needs not completely satisfied by animal products. What, then, is the optimal proportion of animal matter in the diet, and what factors influence this proportion? Since no omnivorous primate subsists even for a day solely on animal matter, there may never be a conflict between animal matter consumption and nutritional adequacy. Nor is there any evidence to suggest that consumption of various classes of dietary constituents by freeranging primates has any adverse effects, although this generalization may be made more from ignorance than from quantified measurement. On the other


4. Omnivorous Primate Diets

hand, there is evidence that secondary compounds may reduce the digestibility of certain plants (e. g., McKay et al. 1978). Optimization of energy gain needs to be evaluated relative to the costs of ingestion of deleterious foods, and other foraging costs. Table 4.1 summarizes the bulk percentage of various dietary categories consumed by five chacma baboon troops representing two populations. These data show that items other than animal matter make up the bulk of the annual diet. 13ut these data alone say little or nothing about preference. Often there are no fruits and seeds, and only traces of animal matter, readily available to these animals. As the environment changes throughout the year and between years, preferences can be identified only by observing choices when alternatives are present. Two major factors determine these choices: the availability of alternative kinds of food, and the amount of effort required to process a food type and obtain its desirable or usable portion. Additional evidence comes from the daily intake pattern for food classes. Often, available animal matter and fruits are eaten in the first hours of the day. Individuals may then become less selective throughout the day. We (Hamilton, Buskirk, and Buskirk 1978) frequently observed adult male chacma baboons eating handfuls of the relatively unpalatable silicaceous leaves of Salvadora persica late in the day, but never in the first half of the daily cycle. Based upon our observation of chacma baboons in the presence of food alternatives, a clear order of preferences can be determined: 1. Least preferred: leafy material, including grasses 2. More preferred: rhizomes, fresh grass seeds, fruits, and flowers 3. Most preferred: animal matter, including, but not limited to, insects I)erivation of this order is based upon the observation that grasses and their choice basal parts are almost completely ignored when fresh acacia seeds (which are soft and do not need to be cracked) and fruits of various sorts are abundant. When occasional insect outbreaks occur, the troops head for these concentrations in the morning and spend most of the day eating them, bypassing the usual seeds and fruits (Hamilton, Buskirk, and Buskirk 1978). At the same time, neighboring troops that do not have the host plants of the erupting insect populations in their home ranges were consuming other, non-animal foods. Nonnally insects become available in small, localized patches. In November 1982 caterpillar larvae were locally infesting the foliage of the euphorbiaceous tree, Croton megalobotrys. For over ten days these sites of insect infestation \vere the first target of the foraging baboons. After an hour or two of foraging in the infested areas, capture rates declined, and the troop moved off to forage on the bases of fresh grass stalks. During the following days they spent their rnornings digging in shallow duff, manually ripping open cocoons and eating


II. Bioevolutionary Antecedents and Constraints

pupae one by one. This, too, was a morning activity, followed by consumption of herbaceous material for the rest of the day. No fruits were available at the time. Sometimes heavy insect outbreaks are distant from roost locations. In the case of the coccid insect outbreak described in Table 4.1, morning movement to this resource took the troop to the most remote part of its home range. Hence, encounter rates and accessibility cannot account for the morning choice of insects in this case. Optimum foraging models based upon random movement through home ranges are thus not applicable. Our focal chacma baboon troops achieved their lowest success in insect matter consumption for termites, solitary grasshoppers, and ants. Consumption of these insects is an individual and facultative activity, and although there is intense competition for these sometimes highly localized resources, they provide a relatively small part of individual baboon diets.

Tooth Micro-wear Patterns
Considerable recent attention has been given to the micro-wear patterns of australopithecine and other mammalian teeth in an attempt to establish correlations with diets. On the basis of tooth micro-wear patterns, Walker (1981) concludes that the robust australopithecines were neither grazers nor browsers and that they did not eat husked (hard) seeds. He therefore dismisses Jolly's (1970) australopithecine seed-eating hypothesis and Szalay's (1975) suggestion that the heavy enamel of australopithecine teeth is an adaptation to bone-crushing or -crunching, since both seed-cracking and bone-crunching produce micro-wear characteristics not found on australopithecine teeth. His further conclusion, that australopithecines were frugivores, is based upon his observation that the tooth micro-wear patterns of 20 East African specimens cannot be distinguished from those of drills, chimpanzees, and orangutans, which he assumes to be frugivores. Walker (1981) further assumes that the seeds of fruits eaten by australopithecines were not processed, as this also would have produced distinctive microwear patterns. There is much to say about the emerging evidence from this promising avenue of research, and what studies of contemporary primates and their diets may contribute. I know of no measurements of baboon tooth micro-wear, yet much of my subsequent discussion of primate diets focuses upon them. Baboons do process fruits without, in most cases, attempting to crush their seeds. Fleshy portions of stone fruits are usually removed by manual or oral manipulation; less frequently, the seeds are passed. The only persistent crushing I have observed is of dry acacia seeds. Nevertheless, these seeds figure prominently in most savanna and hamadryas baboon diets. Insect chitin, which


4. Omnivorous Primate Diets

also produces distinctive tooth micro-wear patterns, would not cause much tooth abrasion in baboons, since most baboon insect-foraging is for caterpillars, soft insects, pupae of various sorts, and soft-bodied termites. Although I observed one population of baboons feeding almost exclusively on large adult grasshoppers (Schistocerca) for over two months, other troops may for years eat only a few grasshoppers and almost no other hard-bodied, chitinous insects. Thus, opportunism, selection, and processing of food items by omnivorous primates may limit the degree of certainty about what can be concluded from tooth micro-wear patterns. Analysis of tooth micro-wear seems to be a valuable method for excluding certain kinds of food processing, but because of the several ways that some omnivores can process particular food items, microwear studies may have limited utility in determining the kinds of food that are actually eaten. Even in the case of grass, which produces distinctive striations on teeth (Baker, Jones, and Wardrop 1959), tooth micro-wear evidence may not identify the extent to which it is eaten. Baboons usually process grasses so that only soft basal parts are eaten. In the case of vertebrate consumption, baboons are meticulous, rejecting bones and pelts while stripping off and consuming soft flesh. It may be useful not to know what baboon tooth wear patterns are for the moment, so that the evidence of tooth micro-wear can be tested without the risk of post hoc reasoning. Presumably tooth micro-wear patterns are generated by persistent, as opposed to casual, use of particular dietary items. For baboons I would predict some evidence for the consumption of grass, and little else. Striations from grass consumption may be more pronounced for some savanna baboon populations than for others, since there are significant differences between populations in the amount of grass in the diet. ()ther kinds of plant foods, including certain fibrous fruits, may also influence baboon tooth wear patterns. I would predict individual population, age, and sex differences in baboon tooth micro-wear patterns, with more variation than has been observed for mammals with more stereotyped diets. Our current knowledge of baboon and other omnivorous primate diets suggests that there may also be considerable population variation in australopithecine tooth micro-wear patterns, if australopithecines were omnivores too. Thus, Walker's (1981) identification of possible australopithecine diets may need to be enlarged to include a much more diverse diet.

IIunter-Gatherers as Models
Most populations of human hunter-gatherers were exterminated by the spread of other cultures before their diets were well studied. Analysis of the diets of extant hunter-gatherers, while interesting, is in part compromised by changes


II. Bioevolutionary Antecedents and Constraints

in their densities compared with aboriginal conditions, and by culturally induced changes in diet. Interpretations are further complicated by omnipresent exchange systems between tribes, often involving food. The hunter-gatherer populations that lived on the richer soils of the planet were replaced long ago by agriculturalists. Since rich soils support rich vegetation, and a potentially greater biomass of non-human animal species, it follows that pre-agricultural humans probably lived at their greatest densities in these areas. Now the only relatively undisturbed hunter-gatherers live in arid regions or other areas not well suited to agriculture and pastoralism. Studies of the diets of hunter-gatherers now living on marginal lands probably provide poor and incomplete information about the diet of the vast majority of early humans. An ecologist might insist upon the following characteristics before considering a population appropriate for the evolutionary analysis of diet: 1. Mean density determined by evolutionarily relevant parameters-not by human disturbance other than that of the subject population in the case of hunter-gatherers 2. Dietary items that are the same as or equivalent to those which shaped the dietary habits of the species 3. Social groupings that have approximately the character and size of evolving populations (i. e., they are not modified by the provision of human agricultural products or freed by human interference from density-regulating processes by predators)
These are not difficult requirements to fulfill for many small animals, or even for some larger species living in areas remote from human activities. They are also relatively easily met by dietary specialist species. Problems arise in the data base and its interpretation for those omnivores whose diets broadly overlap those of humans and who live in the vicinity of human populations. It is not only humans whose diets have been modified by agriculture and other forms of human-induced change. Baboons in particular-at least the majority of populations that live near human habitation-also accept the easy calorie. In the case of extant hunter-gatherers, the intrusion of trade has changed the availability of food via the introduction of new items. Improvisations such as spears, bows and arrows, gill nets, guns, and dogs have also modified the hunting process. A cultural anthropologist would be fortunate indeed to find a population of hunter-gatherers uninfluenced by any of these extra-cultural artifacts and conditions. Another serious interpretive dilemma concerns density. Ecologists assume that over a short period of perhaps a few millennia or a few centuries, the density of local animal populations fluctuates around some relatively constant mean. This has many implications for determinations of a "natural" diet. Over a few years one can, with some luck, witness some years of extreme weather


4. Omnivorous Primate Diets

and expect conditions for the study population to approximate the evolutionarily relevant diet and conditions stated in the foregoing list. Schaller and Lowther (1969) attempted to determine the ability of early hominids to make a living scavenging on the East African plains. To do so, they strolled across the Serengeti plains, looking for partially eaten and abandoned lnammal carcasses, helpless gazelle fawns, and other easily captured prey. Although I admire the direct approach they took to the problem, I am unconvinced by their conclusions. The experience of two men scavenging in a vast park not otherwise hunted by humans does not satisfy conditions 1 and 3 of the list. If hunting success were density-dependent, Schaller and Lowther's (1969) Serengeti experience might have been quite different. Persistent scavenge hunting (Hamilton 1973) by larger groups of hominids over long intervals might not ha\;~ been as successful. Prey availability is limited not only by prey abundance OLIt also by prey wariness, which increases with hunting pressure. A more appropriate but impractical experiment would be to introduce social groups to an area and to measure their foraging success over a period of years, observing the eventual equilibrium between human density and prey- and carrion-procurement success.

Optimality Models
()ptimality models for foraging predict the outcome of an animal's choosing foods that provide an adequate diet at the least cost and risk. Optimality models emphasize the time and energy costs of handling (processing) food, the risks entailed in doing so, and the need to obtain a balanced diet. Under the basic assumption that evolution has produced maximally efficient design (Krebs and Davies 1981), the foraging behavior of an organism or population closely Inatches changed environments. Thus, "legacy" behaviors discussed earlier are outcomes not predicted by optimality models, since they are dietary preferences that have not kept pace with environmental changes. This does not mean that if we accept a particular legacy argument we must reject the applicability of optimality theory to human dietary choices. It suggests only that there may be cases where non-optimality can be identified, perhaps because of the rapidity of environmental change. For contemporary humans in particular, a simple optimality model for choice of food seems unlikely to be applicable, since culture has probably not closely tracked natural selection. It is possible to argue that humans evolved dietary preferences that, in the past, optimized the quality and quantity of their diet, and that the contemporary mismatch of diet to nutritive needs is based upon prehistoric evolutionary events.


II. Bioevolutionary Antecedents and Constraints

Until there is a more unified and widely accepted set of hypotheses concerning what culture is and how it influences human dietary habits, an evolutionary theory of human food choices will not be possible. The position taken in this paper is that culture adapted early hominids to the environment more precisely than it now does for most contemporary human populations. This suggests that human culture has produced evolved characteristics that today lead some contemporary populations, indirectly or directly, to adopt diets not suited to optimum health. If a degree of fixation of choice for broad categories of food items is the basis of some human food preferences, it may be possible to find primate or animal species that have evolved similar fixations. Certain living omnivorous primate species may be well suited to analyses relating food availability to the extent to which it is eaten. If the assumptions stated in this paper are valid, studies of the diets of omnivorous primates can tell us much about why many contemporary humans eat the foods they do, regardless of whether they are the best available diets to sustain optimum health.

References Cited
Alexander, R. D. 1971 The Search for an Evolutionary Philosophy of Man. Proceedings of the Royal Society of Victoria, 84: 99-119. Alexander, R. D.; ]. L. Hoagland; R. D. Howard; K. M. Noonan; and P. W. Sherman 1979 Sexual Dimorphisms and Breeding Systems in Pinnipeds, Ungulates, Primates, and Humans. In Evolutionary Biology and Human Social Behavior, N. A. Chagnon and W. Irons, eds., pp. 402-35. North Scituate: Duxbury. Baker, G.; L. H. P. Jones; and I. D. Wardrop 1959 Causes of Wear on Sheep's Teeth. Nature 184: 1583-85. Balling, ]. D., and]. H. Falk 1982 Development of Visual Preference for Natural Environments. E nvironmen! and Behavior 14:5-28. Bigelow, R. 1969 The Dawn Warriors. Boston: Little, Brown. Boyd, R., and P. ]. Richerson 1982 Cultural Transmission and the Evolution of Cooperative Behavior. Human Ecology 10:325-51. 1985 Culture and the Evolutionary Process. Chicago: University of Chicago Press.


4. Omnivorous Primate Diets

Cook, D. C., and j. E. Buikstra 1973 Circular Caries: A New Tool in Nutritional Assessment in the Past. Paper read at the annual meeting of the American Association of Physical Anthropologists, Dallas. Durham, W. H. 1978 Toward a Coevolutionary View of Human Biology and Culture. In The Sociobiology Debate, A. Caplan, ed., pp. 428-48. New York: Harper and Row. Eibl-Eibesfeldt, 1. 1979 The Biology ofPeace and War. New York: Viking. Fischler, C. Food Preferences, Nutritional Wisdom and Sociocultural Evolution. In 1981 Food, Nutrition and Evolution, D. N. Walcher and N. Kretchmer, eds., pp. 59-68. New York: Masson. Hamilton, W. ]. 1973 Life's Color Code. New York: McGraw-Hill. Hamilton, W. ].; R. E. Buskirk; and W. H. Buskirk 1978 Omnivory and Utilization of Food Resources by Chacma Baboons, Papio ursinus. American Naturalist 112:911-24. Hamilton, W. ]., and C. D. Busse 1978 Primate Carnivory and Its Significance to Human Diets. Bioscience

Social Dominance and Predatory Behavior of Chacma Baboons. Journal ofHuman Evolution 11: 567-73. Isaac, D. L., and D. C. Crader 1981 To What Extent Were Early Hominids Carnivorous? An Archaeological Perspective. In Omnivorous Primates, R. S. O. Harding and G. Teleki, eds., pp. 37-103. New York: Columbia University Press. Jolly, C. ]. 1970 The Seed-Eaters: A New Model of Human Differentiation Based on a Baboon Analogy. Man 5:5-26. Kay, R. F. 1984 On the Use of Anatomical Features to Infer Foraging Behavior in Extinct Primates. In Adaptation for Foraging in Nonhuman Primates, ]. G. H. Cant and P. S. Rodman, eds., pp. 21-53. New York: Columbia University Press. Kay, R. F., and W. L. Hylander 1978 The Dental Structure of Mammalian Folivores with Special Reference to Primates and Phalangeroidea (Marsupialia). In The Ecology of Arboreal Folivores, G. G. Montgomery, ed., pp. 173-196. Washington, D. C.: Smithsonian Press. Krebs, j. R., and N. B. Davies 1981 An Introduction to Behavioral Ecology. Sunderland, Mass.: Sinauer Associates.



II. Bioevolutionary Antecedents and Constraints

Kretchmer, N. 1972 Lactose and Lactase. Scientific American 227: 71-78. 1981 Food: A Selection Agent in Evolution. In Food, Nutrition and Evolution, D. N. Walcher and N. Kretchmer, eds., pp. 37-48. New York: Masson. Lumsden, C. ]., and E. O. Wilson 1981 Genes, Mind and Culture: The Coevolutionary Process. Cambridge: Harvard University Press. McKay, D.; P. G. Watterman; C. N. Mbi; ]. S. Cartlan; and T. T. Strutsaker 1978 Phenolic Content of Vegetation in Two African Rain Forests: Ecological Implication. Science 202: 61-64. McNicholl, B., et al. 1981 History, Genetics and Natural History of Celiac Disease-Gluten Enteropathy. In Food, Nutrition and Evolution, D. N. Walcher and N. Kretchmer, eds., pp. 169-77. New York: Masson. Orians, G. H. Habitat Selection: General Theory and Applications to Human Behavior. 1980 In The Evolution ofHuman Social Behavior, ]. S. Lockard, ed., pp. 4966. New York: Elsevier. Pulliam, R. H. 1974 On the Theory of Optimal Diets. American Naturalist 108: 59-74. Pyke, G. H.; H. R. Pulliam; and E. L. Chamov 1977 Optimal Foraging: A Selective Review of Theory and Tests. Quarterly Review ofBiology 52: 137-54. Schaller, G. W., and G. R. Lowther 1969 The Relevance of Carnivore Behavior to the Study of Early Hominids. Southwestern Journal ofAnthropology 25:307-4l. Schoener, T. W. 1971 Theory of Feeding Strategies. Annual Review of Ecology and Systematics. 2:369-404. Simoons, F. ]. Primary Adult Lactose Intolerance and the Milk-Drinking Habit: A Prob1970 lem in Biological and Cultural Interrelations, II: A Cultural-Historical Hypothesis. American Journal ofDigestive Diseases 15:695-710. Celiac Disease as a Geographic Problem. In Food, Nutrition and E volu1981 tion, D. N. Walcher and N. Kretchmer, eds., pp. 179-99. New York: Masson. Szalay, F. S. Hunting-Scavenging Protohominids: A Model of Hominid Origins. Man 1975 (n. s.) 10:420-29. Walker, A. Diet and Teeth. Philosophy Transactions of the Royal Society ofLondon B 1981 292:57-64.



Fava Bean Consumption: A Case for the Co-Evolution of Genes and Culture
OVER THE LAST 15 YEARS MY COLLEAGUES AND I HAVE BEEN investigating the interface between human nutritional needs and the traditional cultural food practices that satisfy the nutrient needs of the individual and the population as a whole (Katz and Foulks 1970). For example, we have studied the interrelations between human biology and culture in the traditional preparation and processing of maize into specific foods by American native populations (Katz, Hediger, and Valleroy 1974). In this work we have developed evidence that the traditional cultural practices encompassing the preparation and processing of maize significantly enhance its nutritional quality. Maize was the major food supporting the growth of Meso-American civilization. Alkali treatment almost certainly facilitated the intensification of maize agriculture, which in tum led to substantial modifications in the organization and structure of Meso-American society. These social modifications ultimately had effects on the demography and the ecology of Meso-American populations and hence influenced their genetic composition. Thus, over many centuries there was an indirect link between the cultural processes underlying the development of this technology and biological adaptability and evolution within these societies (Katz, Hediger, Valleroy 1975). Although maize preparation practices offer considerable insight into the evolution of cultural practices linked to the supply of essential nutrients, their direct effects on the genetic composition of the population, and hence its biological evolution, have not been amenable to further quantitative analyses. In order to take this biocultural evolutionary approach further, we have been attempting to develop models and hypotheses about how food practices influence specific genetic characteristics. Accordingly, this paper considers and extends our work (Katz 1973, 1979; Katz and Schall 1979) on the cultural factors underlying fava bean consumption and the genetics of favism (a severe hemolytic anemia) in the circum-Mediterranean region. Since the genetic condition under-


II. Bioevolutionary Antecedents and Constraints

lying favism is already documented, the test case presented in this paper allows for the development of more comprehensive heuristic models that use various human foods to evaluate quantitatively the co-evolutionary interactions between biologically based genetic adaptations and traditional cultural practices. The consumption of fava beans has been associated with some of the strongest dietary aversions and prescriptions in recorded history. In fact, it is reasonable to hypothesize that there is a direct biocultural evolutionary connection between the traditional aversion to fava bean consumption and its effects on genetically susceptible individuals. To explore and develop this hypothesis, this paper will briefly review some of the salient aspects of biocultural evolutionary theory and then examine fava bean consumption as a test case of some of the important biocultural, ecological, and evolutionary problems confronting the conceptualization of food preferences and aversions.

Food and Biocultural Evolution
The heuristic biocultural evolutionary approach holds that in every human population there is a time-dependent dynamic equilibrium among ecological, sociocultural, human biological, and demographic variables (see Figure 5.1). The

FIGURE 5.la. The Ecosystems Approach



NOTE: This figure shows the interactions among the following: 1) the biological dimension, which consists of the genetic and phenotypic characteristics of the population; 2) the sociocultural dimension, which consists of the cultural history and the current knowledge or information that is resident in the minds of its members and is accessible to the population as a whole (and includes its material products); 3) the environmental, which includes the non-cultural aspects of the ecosystem in which the population resides; and 4) the demographic dimension, which includes the structure, size, and composition of the population. The interactions are to indicate a dYnamic equilibrium where all of the variables are continuously interacting over time and the information content of the system can be continuously changing.


5. Fava Bean Consumption

mechanisms of evolution that change the gene pool of a population can also be conceptualized as shifting the genetic "infonnation" pool of the population. Cultural traditions, practices, and knowledge fonn a "cultural infonnation pool," which complements and supplements the "biological information pool," and these processes of interaction form a dynamic equilibrium over time in any particular ecosystem. This conceptualization of biocultural evolution assists in the testing of hypotheses about the mechanisms of storage, change, and transmission within the biological and sociocultural information pools and particularly about the interactions between the two pools over a wide range of time intervals (Katz and Schall 1979; see also Katz 1973).

FIGURE 5.1b. Information and Biocultural Evolution




NOTE: The mechanisms underlying the evolution of the genetic information changes are understood relatively well. and the mechanisms by which the human central nervous system receives, processes, creates, stores, and sends its information are being successfully investigated. Likewise, there is increasing systematic understanding about the analogous ways in which the sociocultural system develops, changes, stores, transforms, and communicates (i. e., evolves) its information, particularly in response to the needs of the biological system. However, to systematically integrate the knowledge we have concerning evolution of these three interacting systems (i. e., genetic, neurobiological and behavioral, and sociocultural) represents one of the greatest challenges to anthropology. This paper seeks to explore this problem in presenting an integrated approach to the consumption of fava beans, which represents a particularly rich source of data and hypotheses relevant to this topic.


II. Bioevolutionary Antecedents and Constraints

In the context of this broad biocultural evolutionary model, there are some obvious constraints upon the evolution of implicit and explicit knowledge about foods and their consumption within a society. Understanding the full significance and origins of this knowledge could significantly help in our understanding of some aversive food behaviors. Among other factors, we know that there is a biological component accounting for the "fit" between the food consumed and biological needs and processes. These needs can be divided into various categories of nutritional value (such as protein efficiency and related ratios, caloric content, trace elements, vitamins, and so forth). Obviously this fit between food and nutritional value becomes more significant as the food ascends a scale of importance as a source of sustenance within a particular population. The ecological component is a crucial variable associated with the environmental constraints upon the relative abundance of the food and the ease with which it can be gathered or hunted in the wild state, cultivated, or domesticated. Climatic variables such as rainfall, altitude, sunlight, and temperature, together with soil conditions, all help to establish a possible range of variation. Another component consists of the pharmacological properties of the food. Many edible plants contain highly active toxins, psychoactive drugs, and a wide variety of other compounds whose influence can mimic or even replace the full range of artificially synthesized drugs. Clearly, the special effects of psychoactive drugs synthesized or extracted from plants (e. g., alcohol and coca) must be taken into consideration-for example, their addictive properties. Related to the pharmacological variable is a genetic component, which can limit the fit between populations of consumers and the metabolism and digestibility of the foods they eat. Often the natural constituents of foods contain a factor or factors that may be specifically compatible or incompatible with the genetic constitutions of individuals within the population (Katz 1979). These biological factors play roles of varying importance in the development and evolution of cultural "knowledge" concerning the processing of food from the raw to a "cooked" state. This leads to several questions about the fit between biological and cultural factors. (1) How close is the fit between the knowledge that accumulates in traditional societies over time and present scientific evidence about optimal patterns of preparation and consumption? For example, to what degree are culturally prescribed food combinations during the same meal or time period nutritionally advantageous? (2) What are the principles by which societies acquire, maintain, and distribute this knowledge of foods over time and space? (3) To what degree is this knowledge about foods linked to other practices within society, and to what extent does it provide an adaptive basis for integrating other aspects of traditional practices, beliefs, myths, and symbols? (4) What are the conditions that favor experimental or nonadaptive uses of foods, whether traditional or nontraditional? (5) What is the nutritional significance of various processing traditions such as heating,


5. Fava Bean Consumption

fennentation (with yeasts, molds, bacteria, etc.), sprouting, peeling, drying, mashing, spicing, and the combinations with other foods and chemicals that are associated with the transformation of raw produce into foods? (6) What governs the degree of explicit (as opposed to implicit) knowledge a society accrues about specific foods or food in general? (7) What role do food aversions play in this evolutionary process? Not only could answers to these and other questions provide the basis for more cogent theoretical developments on the relationship between food and biocultural evolution; they could also have an immediate impact upon our specific understanding of such contemporary human food behaviors as aversions and consumption generally. More theoretically, this kind of conceptualization of the evolutionary feedback relations between and among the elements of the human food chain for any particular population can provide useful heuristic models for the understanding of human ecology. One such conceptualization of the important variables and their relations is presented in Figure 5.2 (see Katz 1982). As our knowledge about any particular food source increases, we can use this kind of heuristic model as a potential resource for fonnulating addi-

FIGURE 5.2. The Human Food Chain

~\ Storage


BtOLOGICr---AL- / - - , /














NOTE: This chain presents a more complex model based on a more detailed development of Figure 5.la. The dimensions are designed to show some of the relations among the important variables in the human food chain. The environmental, the sociocultural, and the biological are all developed more completely to show the continuity of relationships throughout the chain.


II. Bioevolutionary Antecedents and Constraints

tional questions. The aim of the remainder of this paper is to exemplify the utility of this approach for examining the interface between biological evolution and adaptation and the cultural beliefs and practices that shape human food behaviors, using fava beans as an example of a most unusual, important, and often deadly food resource.

The Distribution of Favism
Fava bean consumption provides a particularly interesting test case, since it involves a genetically mediated sensitivity that has been clearly associated in modern times with a potent food aversion in some populations. In the circumMediterranean region, individuals with a sex-linked genetic deficiency of the red blood cell (R B C) enzyme, glucose-6-phosphate dehydrogenase (G 6 P D) deficiency or Gd -, are higWy sensitive to RB C hemolysis after consumption of fava beans (Viciafaba) (Bottini et al. 1970; Katz and Schall 1986; Livingstone 1967, 1973; Motulsky 1965). Usually within several hours, but occasionally up to nine days, after consuming fava beans, hemizygous deficient males and homozygous deficient females can suffer a severe hemolytic anemia (Livingstone 1973; Mager, Razin, and Hershko 1969; Mager, Chevion, and Glaser 1980; Sartori 1971). In approximately 1 in 12 cases in modern times, this anemia can result in death (Belsey 1973; Bottini 1973; Huheey and Martin 1975; Livingstone 1964), and before modern medical treatment was instituted over the last few decades, the mortality rate was probably even greater.

Worldwide Distribution of G6PD Deficiency
Although favism occurs principally in the circum-Mediterranean region (Motulsky 1965), Gd - is found throughout the tropical regions of the world (Bottini et al. 1970; Katz 1982; Katz and Schall 1986), throughout Europe, northern and equatorial Africa, the southern Caspian region, the entire Middle East, India and southeast Asia, southern China (Canton), Indonesia, and New Guinea. Since hemizygous males all have the full potential for developing anemia upon exposure to various foods, drugs, and viruses, it can be accurately stated that G6PD deficiency is the most common known genetic disorder in the world (Katz, Adair, and Schall; Katz and Schall 1977; Katz et al. 1978; Kirkman 1968; Motulsky 1965). Livingstone and a number of other investigators have hypothesized on the basis of its semi-tropical and tropical distribution that G6PD deficiency is associated with resistance to malaria (Bienzle et al. 1972; Livingstone 1967, 1973; Luzzato, Usanga, and Reddy 1969; Siniscalco et al. 1961; Siniscalco et al.


5. Fava Bean Consumption

1966). However, it is important to point out that there are at least as many variants of the enzyme as there are of hemoglobin; hemoglobin variants number well over a hundred (L. Luzzato, pers. comm.; see also Belsey 1973).1 The most common variants are GdB -, found in the circum-Mediterranean area, which has the least enzymatic activity (0- 7 percent), probably accounting for its victims' particular sensitivity to fava bean consumption; GdA - , found in tropical Mrican populations, with moderate activity (8-20 percent); and Gd Canton -, found first in China, also with moderate activity. Elsewhere I have hypothesized that forms other than GdB - interact with naturally occurring dietary oxidants in their particular ecosystem (Katz and Schall, 1986).

Favism and Malaria
The occurrence of GdB - gene in those populations that regularly consume fava beans is quite high, with frequencies between 5 and 30 percent regularly reported in the literature (Livingstone 1967, 1973; Luzzato, Usango, and Reddy 1969; Motulsky 1965; Siniscalco et al. 1966). With this high an occurrence of a serious illness, it is highly likely that extensive biocultural adaptations have taken place over the thousands of years since this gene first evolved in the circum-Mediterranean region. My colleagues and I initially hyopthesized that regardless of the nutritional value and agricultural productivity of fava beans, the mortality rate from favism would be high enough that other crops would have evolved to take their place. Since, however, their consumption has continued essentially unabated, we reasoned that extensive knowledge would evolve about the sensitivity of susceptible individuals in order to minimize the beans' effects (Katz 1979, 1982; Katz, Adair, and Schall 1975; Katz and Schall 1977, 1979, 1986; Katz et al. 1978). Nevertheless, the evidence that favism is still a major disease throughout the region suggests that this evolved knowledge or "cultural information" was not sufficient to overcome the harmful effects of continued consumption (Katz 1979). In a further attempt to resolve this evolutionary paradox, we generated another hypothesis that examines the benefits of the continued consumption of fava beans, particularly in view of the presence of seasonal malaria in most of the region. We were impressed by the fact that highly effective anti-malarial drugs such as primaquine also produce a hemolytic anemia in GdB - (Mediterranean type deficient) individuals. Hence, we reasoned that the mechanism of action of these drugs might be correlated with the effects of fava beans (Huheey and Martin 1975; Katz, Adair, and Schall 1975). In other words, antimalarial drugs that are potent oxidants may stress the GdB- RBCs in the same way that fava beans do. Moreover, if this was the case, it was also possible that this response by GdB - cells accounted for the mechanism by which


II. Bioevolutionary Antecedents and Constraints

FIGURE 5.3. Geographic Distribution of Fava Bean Consumption, Malaria, and G6PD Deficiency





~ Fava bean cultivation

[]IIJ C I G 6 P D deficiency • Co-occurrence of all features

GdB - cells maintained their resistance to malaria. The latter hypothesis suggested furthermore that fava beans could have anti-malarial properties for individuals not susceptible to their serious hemolytic effects. In temlS of biocultural evolution, this hypothesis suggests that a balance exists between the selective advantages of the GdB - gene in areas with high malaria and the advantages of culturally encouraged consumption of fava beans for GdB (normal) individuals in the same areas (see Figure 5.3). We have not yet carried out a definitive field test of the in vitro malarial resistance of RBCs taken from individuals with normal and abnormal genotypes before and after the consumption of fava beans in circum-Mediterranean populations. There is, nevertheless, considerable evidence already available that supports the hyopthesis that fava beans have anti-malarial properties. These data are presented under a series of biochemical and pharmacological, epidemiological, nutritional and agricultural, population genetic, and cultural headings.

Biochemical and Pharmacological Factors
Recently developed biochemical and pharmacological evidence on the mechanism by which an RBC is resistant to malaria suggests that the intra-erythrocytic stage of the malarial parasite life cycle is particularly dependent on


5. Fava Bean Consumption

erythrocytic nicotinamide-adenine dinucleotide phosphate (N AD PH) as a source for maintaining high levels of reduced glutathione (GSH) (Crosby 1956; Eaton and Eckman 1977; Eaton et al. 1976; Eckman and Eaton 1977; Etkin and Eaton 1975; Friedman 1978; Sartori 1971). GSH is probably utilized for maintaining the protein synthesis necessary for this phase of the parasitic life cycle (Kosower and Kosower 1970, 1974). As protein synthesis and other intraparasitic metabolic activities continue, they produce an abundance of hydrogen peroxide, which passes out from the parasite into the R B C. There the hydrogen peroxide is reduced to water and oxygen by a peroxidase enzyme, which requires erythrocytic stores of GSH to function (Liebowitz and Cohen 1968). In the case of the GdB - RBC, the enzyme G6PD has only 1 to 10 percent of its nonnal activity (Eaton et al. 1976). This is critical, since the G6PD enzyme is the source of the glucose-6 phosphate metabolism necessary for N AD PH fonnation, which is used in turn for the parasite metabolism. This condition results in a considerable decrease of available N AD PH for the production of both RBCs and parasitic GSH. Without sufficient GSH to reduce the rapid build-up of hydrogen peroxide (H 2 0 2 ) in the RB C (Liebowitz and Cohen 1968), there is a rapid increase in methemoglobin, Heinz body formation, and subsequent oxidant damage from free radicals upon the membrane of the R B C (Friedman 1978). Several investigators (e. g., Eaton et al. 1976; Friedman 1978; Sartori 1971) have postulated that the added oxidant stress provided by a malarial parasite in a G 6 P D-deficient erythrocyte is sufficient to result in a rapid build-up of oxidant-induced damage to the RBC membrane, so that it either prematurely ruptures or is sequestered by the spleen. (It is possible that the increased sequestration by the spleen and/or premature rupture of the infected RBCs leads to increased anti-body formation against the early fonns of the parasite. This would increase host resistance to the earlier stages of parasitic infection and might provide an efficient mechanism for increased immune resistance to malarial infection; however, this hypothesis has yet to be confirmed.) The net effect is a 2- to 80-fold decrease in the level of malarial infection in individuals with G6PD deficiency (Luzzato, Usanga, and Reddy 1969). Fava beans have a number of strong oxidant pyrimidines (vicine, divicine, isouramil) (Beutler 1970; Chevion and Novak 1983; Jamalian, Aylward, and Hudson 1977; Kosower and Kosower 1967; Lin 1963; Lin and Ling 1962a, 1962b, 1962c). In addition, they contain other, related compounds, as well as high quantities of L-DOPA (Kosower and Kosower 1967), which can be converted either in cooking or in metabolism by tyrosinase to the active oxidant LDO PAquinone (Beutler 1970). Fava beans could, therefore, provide a series of strong oxidant stresses on the GdB- RBC. Experiments in our laboratories (Katz and Schall 1977, 1979) and in others (Walker and Bowman 1960) lead to the conclusion that fava beans dramatically decrease GS H levels and, like anti-malarial drugs, produce sufficient oxidant stress to hemolyse the GdB-


II. Bioevolutionary Antecedents and Constraints

FIGURE 5.4. Effects of Fava Beans and Malarial Parasitism on the Oxidant Sensitivity of the Red Blood Cell


GL UCOS E - - - + - - - - - - - - - - t glucose



G-6-P ~ 6-P-G ~ ~ RIBOSE




FAVA BE ANS-+-------~GSH

~ ~

I ) t PROTE IN SYNTHESIS 2)1' fGSH 3)1' l' H202







H2 0 2

H2 0t0 2



t~rv1EMBR_ANE D~~~~~~

NOTE: This figure appeared in S. H. Katz and]. Schall, Favism and Malaria: A Model of Nutrition and Biocultural Evolution. In N. Etkin, ed., Plants Used in Indigenous Medicine and Dzet: Biobehavwral Approaches. New York: Redgrave Press, 1986. © 1986 by Redgrave Publishing Company. Reproduced here courtesy of Redgrave Press.


5. Fava Bean Consumption

R B C. These findings interrelating glucose metabolism, fava bean effects, vitamin E, the significance of malarial factors, and oxidant stress in the R B Care summarized in Figure 5.4. Furthermore, it is likely that the same oxidant stress occurs in normal, G 6 P D-sufficient cells, except that the stress is not great enough to allow permanent damage from hydrogen peroxide. However, we hypothesize that the combination of oxidant stress provided by consumption of fava beans and infection by the malarial plasmodium is sufficient to interrupt the normal development of the parasite. I presented this concept at the Center for Tropical Diseases in Jerusalem. Subsequent research (Golenser et al. 1983) was designed to determine in vitro the susceptibility of G6PD-deficient and normal erythrocytes incubated with and without isouramil, one of the principle pyrimidine oxidants of fava beans. The addition of isouramil decreased malarial parasite growth rates in G6PDdeficient cells, but not in normal cells. In parasitized erythrocytes, the addition of isouramil had direct anti-malarial effects in both normal and G D P D-deficient cells during the trophozioite and schizont parasite stages (Golenser et al. 1983). This evidence helps to confirm our hypothesis at the biochemical level. Furthermore, I propose that the anti-malarial effect of fava beans is sufficient to create the kind of equilibrium necessary to maintain a balanced presence of the genetic mechanism (G d B -) and a pattern of fava bean consumption in which both factors promote resistance to malarial infections. Additional evidence supporting this hypothesis is the fact that high-level depletion of vitamin E, which normally protects the R B C from the damaging effects of hydrogen peroxide, can result in protection against hemolysis in G d B - individuals (Spielberg et al., 1979). Alternatively, animal studies by Eaton and Eckman (1977) and human studies by Friedman (1978) concluded that vitamin E deficiency allows for a more rapid accumulation of damage to normal cells infected by malarial parasites and a lowering of overall infection rates by intercepting the plasmodial life cycle, presumably in much the same as was in the G6PD-deficient cell. Hence, it is likely that the consumption of fava beans lowers the GSH of GdB+ individuals sufficiently to have a net positive protective effect on their resistance to malaria.

Epidemiological Data
We reasoned that if fava beans had some kind of anti-malarial effect on male hemizygous normals and female homo- and heterozygous individuals, then the fava bean should be widely distributed in association with the occurrence of the G d B - genotype and malaria. The evidence on the distribution of all three variables shows a remarkable concurrence among them, suggesting the likelihood of a significant interaction. Moreover, the available data on the season-


II. Bioevolutionary Antecedents and Constraints

ality of malaria and fava bean consumption in Egypt, Iran, Greece, and Italy (Belsey 1973; Katz, Adair, and Schall 1975) indicate a high degree of overlap between the peak season of consumption and the peak occurrence of the malarial vector (see Figure 5.5 and Katz and Schall 1979). Although these epidemiological data could reflect merely the coincidence between the climatic

FIGURE 5.5. Seasonal Variation in Malarial Vectors, Fava Bean Harvest, and Fava Consumption

Feb. Mar. Apr. May jun. jul.

Aug. sep.




Malana vector








A. pharoensls



Fava harvest Fava consumptIOn

Malana vector




I A. sacharovl I


A. superplctus






Fava harvest Fava consumptIOn

Malana vector

I A. sacharovl


Fava harvest Fava consumptIOn



A. superplctus


Malana vector

A. labranchlal + sacharovl



Fava harvest Fava consumptIOn



NOTE: The shaded areas refer to peak consumption periods for fava beans. The ragged lines separating the species of mosquito indicate overlapping seasonal life cycles for these vectors.


5. Fava Bean Consumption

TABLE 5.1. Comparative Nutritional Values for Mediterranean Region Legumes
% Protein

Chick peas Lentils Kidney beans Fava beans

(dry weight)
17.8 24.0 22.5 23.5

Protein Efficiency Ratio a (cooked legumes)
2.05 ± 0.10 1.14 ± 0.14 1.51 ± 0.22 1.17 ± 0.17

aGrams protein consumed, divided by grams of weight gain, in rats. NOTE: Adapted from FAO Reports on the nutritional contents for legumes (Food and Agricultural Organization, Rome, 1959).

conditions conducive to the spread of malaria and those favoring the cultivation of fava beans, it is unlikely that such a coincidence would be distributed so widely and consistently over the region, particularly if the beans had only negative effects.

Nutritional and Agricultural Potential
Although the epidemiological data show a clear association between the seasonal occurrence of malaria and the consumption of fava beans, this evidence is not conclusive, since the extensive distribution of the fava bean could be explained, for example, by its agricultural and nutritional advantages. In this context it is clear that fava beans are higWy suited to many of the environments in which they are grown. Likewise, their nutritional value is in the same range as that of other regionally grown pulses and legumes, which tend to be relatively high in lysine and low in tryptophan and the sulfur-containing amino acids. Hence, while clearly a higWy productive crop, their presence is not additionally explainable on the basis of any significant nutritional advantages over other pulses and legumes found in these regions. Moreover, these other regional legumes do not contain constituents that trigger hemolytic crises in G6PDdeficient individuals (Table 5.1 and Katz and Schall 1979).

Population Genetics
In order to test the hypothesis that fava beans provide selective advantages for the nonnals of both sexes and heterozygous females and disadvantages for hemizygous deficient males, I carried out analyses of the available genetic data


II. Bioevolutionary Antecedents and Constraints

on population distributions of the GdB - genotype. The usefulness of these rather extensive data is limited by the difficulties in classifying female heterozygotes because of the deactivation of one of the X chromosomes during embryonic development, according to the Lyon hypothesis (Motulsky 1965). The data are also complicated by the fact that other genes for acid phosphatase B (Bottini et al. 1971), thalassemia in females (Carcassi 1974; Friedman 1978), and possibly tyrosinase variation (Beutler 1970) all play roles in mitigating the hemolytic effects of fava beans. Another factor that is likely to be important in the expression of GdB - in females is associated with the onset of adrenarche. Since adrenarche is associated with increased secretion of dehydroepiandrosterone (D H E A), which directly inhibits G 6 P D activity, it is highly likely that the unusual variation in the age at which favism is prevalent in females is closely associated with this developmental event, which occurs at age six to seven (Katz et al. 1983) and is particularly significant in children with higher body mass indexes (Katz et al. 1985). Since these interactions and other unknown factors tend to limit the data, it is more useful to compare the actual ranges of gene frequencies with various hyopthetical outcomes based upon variation in consumption patterns. If we assume that fava bean consumption confers no advantage, then we should see a trend that rapidly proceeds towards a very high frequency of the G6PD-deficiency gene; conversely, if there was a high selective advantage for the consumption of fava beans by normal individuals, the GdB - gene would merely disappear over time. However, if fava bean consumption was somewhat advantageous for the normals and slightly more advantageous for the heterozygotes, which would be the case if fava beans were not consistently available in areas of endemic malaria, then a balanced polymorphism could be maintained in the same range that is widely reported in the circum-Mediterranean region. Hence, the fact that the population genetic data tend to support a balanced genetic polymorphism suggests additional evidence in favor of an anti-malarial effect from fava bean consumption. A more detailed explanation of the population genetics of this phenomenon can be found in Katz and Schall (1986; see Figure 5.6).

The Evolution of Cultural Information
I postulated that if fava beans were consumed over a long enough period of time, the cultural information about them would evolve toward an optimal fit with the cumulative biological effects of the beans. The model would predict a significant growth of cultural mechanisms to prevent the continued consumption of the beans if their net effects were negative. The accumulation of knowledge would proceed to the point where they would no longer be consumed, since danger of death and illness would be correlated with their consumption. On the other hand, if in addition to their negative hemolytic effects on G6PD-


5. Fava Bean Consumption

FIGURE 5.6. Selection Models: Favism, Malaria, and G6PD Deficiency, by Sex




Oth~r R~d C~II



? Cooking




? C!Okon: &


11 Oth~r


G~nes (acId





phosph,bse B.
tyroslnas~ deflcln~cy)

of Carbohydrate

General Health


FE \1-\LE


Norm,ls - - - - -



_------1-------____ t
(G-6·PD qene)



(oth~r red c~1 I

i Lyonlzatlon ...... I
- - - -I I I





I ?Cooklng &

abnormalities plus thalusemla)



(~1 ~ (oth~r r~d c~1I
g~ne\ i



Homozygote DefiCient

?CooktnCJ &






L ~~V~ _I

FAVA (Health Sbtus)

(St,tus of C,rbohydrlt~

NOTE: Males and females are separated to show the differences in selection for this sex-linked gene. The arrows refer to factors that theoretically increase (upward arrow) and decrease (downward arrow) the frequency of the C I G 6 P D Mediterranean gene.

deficient individuals, they had some benefit such as promoting "good health" (i. e., absence of disease due to malaria), we would expect to see a mixture of culturally evolved traditions, both prescriptive and proscriptive, governing the behavior surrounding their consumption. Before the cultural and historical evidence for the pattern of consumption can be evaluated, the time frame of these biocultural evolutionary hyoptheses has to be considered. Fava beans were one of the first plants to be intensively gathered by Indo-European populations. Probably coincident with or anteceding any kind of grain cultivation, they are found in association with neolithic pile


II. Bioevolutionary Antecedents and Constraints

dwelling sites in Switzerland and open-air sites in northern Italy, Spain, and Hungary. Even before their human use in Europe, fava beans were established in a range that extended from Persia and the Himalayas to the East and ultimately to the Atlantic coast in the West. There is other evidence of their use in Bronze age Troy and 12th-Dynasty Egypt, and there are many references to them in classical Greek and Roman literature (Andrews 1949; Arie 1959; Celsus 1935; Giles 1962; Herodotus 1947; Rowlett and Mori 1970). Currently they are known to extend through China and as far as Taiwan. Extensive archaeological and historical data allow for a more careful analysis of our hypothesis, since the period of their continuous cultivation and apparent consumption is so considerable. Given this time dimension, it seems likely that had the effects of fava beans been only negative, their use as a human food would have long since disappeared among populations with G 6 P D-deficiency. Yet it could be argued that no implicit or explicit knowledge ever evolved about their toxicity. This argument, as it applies to the mode of X-linked inheritance (Papavasiliou et al. 1972) of favism, has been made by Giles (1962), who hypothesized that the Indo-European kinship pattern of patrilineality and patrilocality at least tends to preclude the correlation of favism with maternal inheritance in males. That is, the females who carry this X-linked trait to their sons are continually separated out by exogamy from the lineage, which makes it difficult to correlate the illness with family lines. Thus, data about favism could also be used to test hypotheses concerning the ways in which knowledge spreads when knowledge of the disease is carried by women versus men. Presumably women would have observed that their brothers were susceptible to the disease. Therefore, if their sons had the same symptoms as their brothers, it is reasonable to suggest that the women would have drawn some conclusions about the disorder. This knowledge could have given rise to various beliefs and behaviors among women that were different from those of men. A careful investigation of historical data and the roles of women in the transmission of knowledge from one generation to the next under patrilineal, patrilocal, and a partial matrilateral cross-cousin marriage kinship system could yield fruitful tests of hypotheses concerning this aspect of the biocultural evolutionary process. Although favism as a specifically defined disease was first described in the mid-19th century, it is not clear whether it was known by other names or clusters of symptoms at the folk level. For example, through interviews I conducted in areas of Sardinia with very high rates of favism, I uncovered-contrary to the local medical belief-a well-founded folk knowledge of favism that extended back to the childhood of the oldest members of the community, some of whom were over 70. One distinctly recalled being warned about fava beans during childhood by his grandparents. This suggests that knowledge of the beans' negative effects went back well into the 19th century. Nevertheless, it is


5. Fava Bean Consumption

difficult to trace explicit knowledge before the 19th century, and this may also explain why favism was not directly correlated with fava consumption earlier. However, this does not rule out the possibility that earlier prescription or proscription of fava beans could evolve without any explicit knowledge of the relationship between bean consumption and favism. More recently, studies have been conducted to determine whether traditional cooking, processing, and other dietary activities significantly alter the toxicity of the bean for individuals with G6PD deficiency (Belsey 1973; Jamalian 1978; Jamalian, Aylward, and Hudson 1977). In other words, it is reasonable to suggest that certain appropriate processing techniques may have evolved to decrease the possibility of fatalities from favism. Since there are considerable variations in processing, only two of these pathways will be considered here. However, before considering processing per se, it is important to point out that many, but not all, populations tend to consume the beans raw without any processing. In the event of an impending fava crisis, it is reported that the folk medical treatment of Iran is a high "sweet" diet consisting of naturally occurring sweet foods. This is believed to reduce the significance of the crisis. Indeed, there may be some medical rationale for this prescription, since the increased availability of glucose to a G6PD-deficient individual may provide sufficient saturation with glucose-6-phosphate to optimize the remaining G6PD activity, which is involved in the production of NADPH and in tum maintains a necessary level of GSH (Katz and Schall 1986; Katz et al. 1978). The GSH in tum prevents the toxic build-up of cellular peroxides and thus lessens the probability of a fava-induced hemolytic crisis. The first step in processing fava beans is soaking. Most ethnic groups studied do soak their dried beans, and it is likely that soaking is sufficient to trigger the first reactions of germination. Although significant changes in the biochemical constituents could result from this germination process, little is known about its effects upon the toxicity of the beans, except that the toxic constituents are partially water soluble and would be discarded with the soaking water. It is known that the sprouting of soybeans, another legume, rapidly eliminates the anti-trypsin protein activity that prevents insect predation. Since the principal mechanism that the fava bean has evolved to protect its seed is the production of pyrimidines, it is likely that germination would produce a significant change in its oxidant activity. A second widely practiced step in processing consists of removing the skin of the bean. This has been studied by Jamalian and associates Oamalian 1978; Jamalian, Aylward, and Hudson 1977), who report high toxicity for all parts of fresh mature seeds, but higher activity in the seed coats than in the fresh seed flesh, although the overall toxicity was lower in the dried beans. Both major processing steps make a difference in the outcomes, and both are widely used. Jamalian and his associates Oamalian 1978; Jamalian, Aylward,


II. Bioevolutionary Antecedents and Constraints

and Hudson 1977) also report that the GSH levels of the processed beans were lowered by incubation, and that the level remaining was sufficient to avoid spontaneous hemolysis. This raises an important and as yet unanswered question: if the oxidant activity is lowered sufficiently to improve the survival of G6PD-deficient individuals, is it still sufficient to add any protection against malaria in the hemizygous males and homo- and heterozygous females? This question suggests that the model of the relations between fava bean consumption, favism, and the genetic factor requires a separate model for males and females, since males are hemizygous and either have or do not have the gene and the potential for the disease, whereas females have three classes of phenotypes, each with varying susceptibility. If the model also includes the potential interaction with malaria and the effects of fava bean consumption on the genetic frequency of the GdB - gene, then the model begins to take on the characteristics of a more formal quantitative model of genetic change. Elsewhere Katz and Schall (1986) have developed a heuristic model of this phenomenon, which is presented here as Figure 5.6.

Folklore and Fava Bean Consumption
The fact that fava beans have had sufficient time to come to some kind of equilibrium with population groups strongly favors the accumulation of many specific beliefs about their effects. Consistent reports in the classical Greek, Roman, Egyptian, and Indian literature associate them with death (Andrews 1949). Andrews has conducted an intensive study of the ambivalence about fava beans. Although he was unable to conclude which factors besides the flatulence associated with fava bean consumption gave rise to the ambivalence of the belief that the beans were occupied by the souls of the dead, he determined that beliefs about the beans were not associated with some ancient vestige of Indo-European totemism, nor did they stem from some historical accident. The evidence he developed is extensive and relevant to the complex evolution of cultural knowledge about fava beans:
The ancients felt toward beans a mingled respect and dread, a complex of emotions suggested by the Greek term lEQoa, which apparently was generally applied to an object believed to be charged with some supernatural force, contact with which might be either beneficial or harmful. Today we generally call this mysterious power mana in its helpful aspect and taboo in its harmful aspect. Beans belonged in the category of objects possessing both mana and taboo. (Andrews 1949: 277)

Such beliefs in tum led to the notion that fava beans had enormous generative power-a power to be avoided by those members of society whose social


5. Fava Bean Consumption

functions required avoidance of death and sought by those whose functions were closely associated with death. Andrews has also documented that the ambivalence and controversy surrounding the consumption of fava beans are both ancient and widespread. He was able to accumulate evidence from the Greco-Roman era, ancient Egypt, India, and 19th- and 20th-century Mrica. In addition, Rowlett and Mori (1970) have accumulated data for English folklore. These cultural data may be briefly summarized by area and time period. GREECE Despite the strong proscriptions against the beans in classical Greece, there is evidence of their continued use and consumption. The strictest aviodance of fava beans there appears to have been practiced by the Pythagoreans. Andrews relates that "the Pythagorean taboo was so stringent that it extended even to treading down the growing bean vine. According to one account . . . Pythagoras, pursued by Syracusan soldiers, could not bring himself to escape by crossing a field of beans, preferring to let himself be taken and killed. According to another. . . some Pythagoreans, fleeing from hostile soldiers, stopped when they came to a bean field in flower and defended themselves to the death" (1949:276). Bean avoidance extended to other elements of the Greek population as well, although no one at that time seems to have known why the custom existed. i\ristotle reported that the Pythagoreans, who were partisans of the oligarchical regime, detested the beans because they were used to cast a vote in elections for magistrates and were therefore a symbol of democracy. Nevertheless, Andrews notes that Aristotle supplied four more explanations for fava bean avoidance without settling on anyone of them. Plutarch also offered various explanations, as did Lydus. Arie (1959) hypothesizes that at the root of the Pythagorean aversion was Pythagoras's own G6PD deficiency and possibly some knowledge about favism. What is evident is that although many ancient Greeks continued to eat fava beans, others shunned them, and those that avoided them had a plethora of reasons for doing so. ROME Judging from such information as Andrews provides, the prevailing rationale for fava bean avoidance among the ancient Romans was that the beans caused bad dreams. For this reason, Pliny, Diogenes Laertius, and Amphiaraus, a mythical dream interpreter, suggested abstention (1949:285). Plutarch noted that consumption of the beans was associated with increased male sexuality, restlessness, dreams, and flatulence. At the same time, others (particularly physicians) felt that fava beans were a good, healthful food. They were popular with artisans, farmers, builders, and gladiators (1949:281n).


II. Bioevolutionary Antecedents and Constraints

ENGLAND Rowlett and Mori note the presence of "unmistakable references to the magical effects of fava beans-or at least their deleterious ones"-in English folklore (1970: 100). In many bean stories and superstitions, of course, it is nearly impossible to determine the identity of the bean or beans involved; but there are some stories in which the involvement of fava beans is quite beyond doubt. One of these is Ralph of Coggershall's tale of the Green Children, in which children eat beans, turn green (which is diagnostic of favism), and die. Similarly, the identification of fava beans in the Jack and the Beanstalk story is certain, as the beans in the story are very large and the stalk is tall and strong. Other folk conceptions concerning fava beans in England include the popular idea that more cases of lunacy occur when the plant is in bloom than at any other time of the year, a notion among coal miners that accidents in the pits occur more frequently when the bean plants are in bloom, and a general reluctance to have bean blossoms in the house. There is also a saying: "Sleep in a bean field all night if you want to have awful dreams or go crazy" (Rowlett and Mori 1970: 101), as well as the idea that eating fava beans can harm one's senses of smell and vision and can be harmful to the blood. Rowlett and Mori suggest that "some of the most persistent English folktales" deal principally with the effects of Vicia Java (ibid.). ANCIENT EGYPT Andrews (1949:277n) notes the existence of bean avoidance in Egypt, but it is not clear how extensive or stringent the taboo was. Herodotus states that Egyptians would neither grow these beans nor eat them raw or cooked, and that they were so strongly proscribed for the priests that they could not even tolerate the sight of them; yet archaeological evidence and papyrus records at least as far back as the 12th Dynasty suggest their widespread and long-term use as a food and in funerary rites. "There is no doubt," says Andrews, "that a taboo of some kind did exist, and various implausible explanations have been advanced" (ibid.). Here again, we seem to have a situation in which fava bean taboos and widespread fava bean consumption exist side by side in the same society. Currently, fava beans are a major staple food for the vast majority of the Egyptian population. INDIA Other evidence for fava bean avoidance comes from Vedic India and the ethnographic present: "A bean taboo imposed on those rendering sacrifice is mentioned in the oldest Indian ritualistic text, the Yajurveda, as well as in the


5. Fava Bean Consumption

Maitrayani Samhita and the Kathaka. The Bagdis of central and western Bengal are apparently of Dravidian descent. In the territory of Bankura, where the original caste structure seems to be particularly well preserved, they are divided into nine endogamous subcastes, which in tum are subdivided into exogamous clans or septs. Many of the latter are totemic, for example, the Patrischi or Bean Clan, members of which will not touch beans" (Andrews 1949: 277n). Once again, it is my experience that fava beans are still widely consumed in western India today. AFRICA Andrews has one short but noteworthy comment on Mrica: "Among the Baganda in the vicinity of Lake Victoria Nyanza, the Bean Clan will not eat or even cultivate beans. One of them is said to have eaten beans and died on the spot" (ibid.).

Other Biological Effects of Fava Beans
Fava beans have a considerable concentration of indigestible oligosaccharides, which are digested by bacteria of the lower gastrointestinal tract, the likely cause of the flatulence reported in the cultural data presented above. However, they also contain a high concentration of L- D 0 PA (approximately 0.25 percent by weight). L- DO PAis a potent psychoactive neurotransmitter that is known to be associated with the other symptoms Plutarch described and is consistent with English folklore. Moreover, the dose of L- DO PA obtained from one or two meals of fava beans appears to be in the range that can be effective in treating Parkinson's disease, which is the principal source of clinical data on the effects of L- DO P A Oamalian 1978; ]amalian, Aylward, and Hudson 1977). Thus, it is conceivable that beside their agricultural efficiency and nutritional value, their possible psychoactive effects may be associated with their continued consumption. This combination of effects appears to fit well with the Greek word hieros, which Andrews (1949) suggested conveys the mixture of dread and respect that was applied to the hannful or beneficial supernatural force associated with the consumption of fava beans. Andrews sums up the magnitude and complexity of the "infonnation pool" that was accumulated in ancient and historic times by concluding that "no plant or animal known to the lndo-Europeans produced a more luxuriant growth of beliefs than fava beans" (1949: 290).


II. Bioevolutionary Antecedents and Constraints

The evidence suggests that fava bean consumption throughout the circumMediterranean region fits a complex model of interacting biological and cultural evolution. The beans undoubtedly have highly toxic effects for many G6PDdeficient individuals, and hence there is extensive evidence for the development of taboos surrounding their use, particularly in children. In addition, many recipes for preparing the beans lower their toxic effects. However, consumption of the beans has continued since neolithic times, and it is clear that neither the taboos nor the processing techniques have mitigated this problem sufficiently to prevent a high incidence of favism. Even though their agricultural potential in the circum-Mediterranean region is high, it is still difficult to explain their continued use in light of the high morbidity and mortality rates in G6PDdeficient individuals. Hence, the evidence tends to support the addition of increased resistance to malaria as another major factor adding selective advantage to their continued consumption. Several classes of pharmacologically active compounds in fava beans that appear to be responsible for the favism crisis in G6PD-deficient individuals produce similar increases in the oxidant sensitivity of the RB Cs in normal individuals without toxic effects. Various epidemiological and biochemical evidence supports the hypothesis that an increase in RBC oxidant sensitivity as a result of fava bean consumption could lower rates of infection with malarial parasites and, therefore, be highly advantageous (Papavasiliou et al. 1972). If this hypothesis is fully substantiated, it would provide the first evidence that the biological and cultural evolution of disease resistance are linked through dietary practices. This paper has also attempted to demonstrate the complex interactions between changes in the gene pool and changes in cultural knowledge, which accumulates over time a set of adaptive practices and information about the effects of fava bean consumption. Topics that call for further exploration in light of these cultural data include quantitative aspects of the evolution of the Mediterranean type of G6PD deficiency; folk knowledge of the sex-specific effects of the beans on males but not females, and the folk behavior and beliefs about females as carriers of the potentially lethal condition; and further research concerning the efficiency of folk medical practices for treating favism and the relations of these practices to the anti-malarial potential of the beans. The mechanism of transfer of this folk knowledge from one generation to the next and the roles of rituals and ideologies in this process could provide important clues for the assembly of a more sophisticated model. Finally, fava beans are still very widely consumed, and there is a need to integrate these findings with knowledge of contemporary social, agricultural, and food practices. The apparent net effect of these interactions is a dynamic equilibrium between biological and


5. Fava Bean Consumption

sociocultural dimensions, which, I suggest, provides a test case for understanding the co-evolution of genes and culture as it relates to the human food chain.

'fhe ideas and data presented in this revised and updated paper are largely derived from Katz (1979). A more extensive discussion of the human biological aspects of the problem is given in Katz and Schall (1986), in which Figure 5.4 appeared. Figures 5.1b, 5.2, 5.3, and 5.6 appear courtesy of AVI Publishing, from Katz: Food, Behavior and Biocultural Evolution, in L. M. Barker, ed., The Psychobiology ofHuman Food Selection, pp. 171-88, © 1982 by the AVI Publishing Company, Westport, CT 06881. Figure 5.5 and 'fable 5.1 appear courtesy of Medical Anthropology (see Katz and Schall 1979), © 1979 Redgrave Publishing Co., Bedford Hills, N. Y.

1. This wide range of enzyme variants has given rise to several notation systems for

C; 6 P D deficiency, such as Gd -, which refers to the deficient variant of the enzyme.

References Cited
Andrews, A. C. 1949 The Bean and Indo-European Totemism. American Anthropologist 51:274-92. Arie, T. H. D. 1959 Pythagorus and Beans. Oxford Medical School Gazette 2: 75-81. Belsey, M. A. 1973 The Epidemiology of Favism. Bulletin of the World Health Organization 48: 1-13. Beutler, E. 1970 L-Dopa and Favism. Blood 36:523-25. Bienzle, U.; A. O. Lucas; O. Ayemi; and L. Luzzato G6PD and Malaria. Lancet 1: 107-10. 1972 Bottini, E. 1973 Favism: Current Problems and Investigations. Journal ofMedical Genetics 10: 213-19. Bottini, E., et a1. 1970 Presence in Vicia faba of Different Substances with Activity in vitro on Gd - Med Red Blood Cell Reduced Glutathione. Clinica Chemica Acta 30:831-34.


II. Bioevolutionary Antecedents and Constraints

Bottini, E., et al. 1971 Favism: Association with Erythrocyte Acid Phosphotase Phenotype. Science 171: 409-11. Carcassi, L. E. F. 1974 The Interaction Between B-Thalassemia, G6 P D Deficiency and Favism. Annals of the New York Academy of Sciences 232: 297-305. Celsus 1935 De Medicina, W. G. Spencer, trans. Cambridge: Harvard University Press. Chevion, M., and T. Novak 1983 Favism. Annals ofBiochemistry 128: 152-53. Crosby, W. H. 1956 Favism in Sardinia. Blood 11:91-92. Eaton, ]. W., and]. R. Eckman 1977 Glutathione Metabolism in Malaria Infected Erythrocytes. Clinical Research 25:610. 1979 Plasmodial Glutathione Metabolism: Dependence Upon the Host Cell. Nature 278: 754-56. Eaton,]. W.;]. R. Eckman; E. Berger; and H. S. Jacob 1976 Suppression of Malaria Infection by Oxidant Sensitive Host Erythrocytes. Nature 264: 75~-60. Etkin, N. L., and]. W. Eaton 1975 Malaria Induced Erythrocyte Oxidant Sensitivity. In Erythrocyte Structure and Function, G. ]. Brewer, ed., pp. 219-32. New York: Liss. Friedman, ]. M. 1978 Increased Oxidant Sensitivity of Malaria Parasites Glucose-6-Phosphate Dehydrogenase Deficient and Thalassemia Trait Red Cells. Blood 52 (suppl. 1): 64. Giles, E. 1962 Favism, Sex Linkage, and the Indo-European Kinship System. Southwestern Journal ofAnthropology 18: 286-90. Golenser, ].; ]. Miller; D. T. Spiro; T. Novak; and M. Chevion 1983 Inhibitory Effect of a Fava Bean Component on the In Vitro Development of Plasmodium falciparum in Normal and Glucose-6-Phosphate Dehydrogenase Deficient Erythrocytes. Blood 61:507-10. Herodotus 1947 The Persian Wars. New York: Random House. Huheey, ]. E., and D. L. Martin 1975 Malaria, Favism, and G6PD Deficiency. Experientia 30: 1145-47. Iversen, L. I. 1975 Dopamine Receptors in the Brain. Science 188: 1084-89. Jamalian, ]. 1978 Favism-Inducing Toxins in Broad Beans (Vicia faba) Determination of Vicine Content and Investigation of Other Nonprotein Nitrogenous Compounds in Different Broad Bean Cultivars. Journal of the Science of Food and Agriculture 29: 136-40.


5. Fava Bean Consumption

Jamalian, ].; F. Aylward; and B. ]. F. Hudson 1977a Favism-Inducing Toxins in Broad Beans (Vicia faba): Biological Activities of Broad Bean Extracts in Favism Sensitive Subjects. Plant Foods for Human Nutrition 27:213-19. 1977b Favism-Inducing Toxins in Broad Beans (Vicia faba): Estimation of the Vicine Contents of Broad Bean and Other Legume Samples. Plant Foods for Human Nutrition 27:207-11. Katz, S. H. Evolutionary Perspectives on Purpose and Man. Symposium on Human 1973 Purpose. Zygon 8:325fl-40. Un Exemple d'Evolution Bioculturelle: La Feve. Communication 31:531979 69. Favism, G6PD Deficiency and Malaria: The Evolution of G6PD Defi1981 ciency. Unpublished Manuscript. Food, Behavior and Biocultural Evolution. In The Psychobiology of 1982 Human Food Selection, L. M. Barker, ed., pp. 171-88. Westport, Conn.: AVI. Katz, S. H.; L. Adair; and J. Schall 1975 Fava Bean Consumption, Malaria and G6PD Deficiency. American Journal ofPhysical Anthropology 44: 189. Katz, S. H., and E. F. Foulks 1970 Calcium Homeostasis and Behavioral Disorders. In Symposium on Human Adaptation, S. H. Katz, ed. Journal of Physical Anthropology 32: 225-316. Katz, S. H.; M. L. Hediger; J. Schall; and L. Valleroy 1983 Growth and Blood Pressure. In Clinical Approaches to High Blood Pressure in the Young, T. and M. Kothchen, eds., pp. 91-132. Boston: John Wright/ PSG Inc. Katz, S. H.; M. L. Hediger; and L. Valleroy 1974 Traditional Maize Processing Techniques in the New World: Anthropological and Nutritional Significance. Science 184: 765-73. 1975 The Anthropological and Nutritional Significance of Traditional Maize Processing Techniques in the New World. In Symposium on Biosocial Interrelations in Population Adaptation, Wayne State University, 1973. In Biosocial Interrelations in Population Adaptation, E. S. Watts, F. E. Johnson, and G. W. Lasker, eds., pp. 195-234. The Hague: Mouton. Katz, S. H., and J. Schall 1977 Fava Bean Consumption, Malaria and G6PD Deficiency. American Journal ofPhysical Anthropology 46: 178. Fava Bean Consumption and Biocultural Evolution. Medical Anthropol1979 ogy 3: 459- 76. 1986 Favism and Malaria: A Model of Nutrition and Biocultural Evolution. In N. Etkin, ed., Plants Used in Indigenous Medicine and Diet: Biobehavioral Approaches. New York: Redgrave Press. Katz, S. H.; J. Schall; P. Sundick; and J. Coleman 1978 Fava, Biocultural Evolution and Favism. American Anthropological Association Abstracts, no. 78.


II. Bioevolutionary Antecedents and Constraints

Katz, S. H., et a1. 1985 Adrenal Androgens, Body Fat and Advanced Skeletal Age in Puberty: New Evidence for the Relations of Adrenarche and Gonadarche in Males. Human Biology 57:401-403. Kirkman, H. N. 1968 G6PD Variants and Drug Induced Hemolysis. Annals of the New York Academy ofSciences 151: 753-64. Kosower, N. S., and E. M. Kosower 1967 Does 3,4-dihydroxyphenyalanine Play a Part in Favism? Nature 215:285-86. 1970 Molecular Basis for Selective Advantage of G 6 P D Deficient Individuals Exposed to Malaria. Lancet 2: 1343-45. 1974 Effect of Oxidized Glutathione on Protein Synthesis. In Genetic Polymorphisms and Diseases in Man, B. Ramot, ed., pp. 349-57. New York: Academic Press. Liebowitz, ]., and G. Cohen 1968 Increased Hydrogen Peroxide Levels in Glucose Erythrocytes Exposed to Acetylphenylhydrazine. Biochemical Pharmacology 17:983. Lin,]. Y. 1963 Studies on Favism, 4: Reactions of Vicine and Divicine with Sulfhydrl Group of Glutathione and Cytesine. Journal of the Formosan Medical Association 62: 777-8l. Lin, ]. Y., and K. H. Ling 1962a Studies on Favism, 1: Isolation of an Active Principle from Fava Beans (Vicia faba). Journal of the Formosan Medical Association 61:484-89. 1962b Studies on Favism, 2: Studies on the Physiological Activities of Vicine in vivo. Journal of the Formosan Medical Association 61:490-94. 1962c Studies on Favism, 3: Studies on the Physiological Activities of Vicine in vitro. Journal of the Formosan Medical Association 61: 579-83. Livingston, F. B. 1964 Aspects of the Population Dynamics of the Abnormal Hemoglobin and G6PD Deficient Genes. American Journal ofHuman Genetics 16:43550. 1967 Abnormal Hemoglobins in Human Populations. Chicago: AIdine. 1973 Data on the Abnormal Hemoglobins and G6PD Deficiency in Human Populations. University of Michigan Technical Reports, no. 3, pp. 1216. Ann Arbor: University of Michigan. Luzzatto, L.; E. Usanga; and S. Reddy 1969 Glucose 6 Phosphate Dehydrogenase Deficient Red Cells: Resistance to Infection by Malarial Parasites. Science 164:839-42. Mager, ].; M. Chevion; and G. Glaser 1980 Favism. In Toxic Constituents ofPlant Foodstuffs, I. E. Liener, ed., pp. 265-94. 2nd ed. New York: Academic Press. Mager, ].; A. Razin; and A. Hershko 1969 Favism. In Toxic Constituents ofPlant Foodstuffs, I. E. Liener, ed., pp. 293-318. New York: Academic Press.


5. Fava Bean Consumption

Motulsky, A. G. 1965 Theoretical and Clinical Problems of Glucose-6-Phosphate Dehydrogenase Deficiency. In Abnormal H aemoglobins in Africa: A Symposium Organized by the Council for International Organizations of Medical Sciences Established Under the Joint Auspices of UNESCO & WHO, ]. H. P. Jonxis, ed., pp. 143-96. Philadelphia: Davis. Mourant, A. E.; A. C. Kopec; and K. Domaniewska-Sobczack 1976 The Distribution of Human Blood Groups and Other Polymorphisms. London: Oxford University Press. Papavasiliou, P. S.; G. C. Cotzias; S. E. Dueby; A.]. Steck; C. Fehling; and M. A. Bell 1972 Levodopa in Parkinsonism: Potentiation of Central Effects with a Peripheral Inhibitor. New EnglandJournal ofMedicine 285:8-14. Rowlett, R. M., and j. Mori 1970 The Fava Bean in English Folklore. Ethnologia Europea 4:98-102. Sartori, E. 1971 On the Pathogenesis of Favism. Journal of Medical Genetics 8:462-67. Siniscalco, M.; L. Bernini; G. Fillipi; B. Latte; P. Meera Khan; S. Piomelli; and M. Rattazzi 1966 Population Genetics of Hemoglobin Variants, Thalassemia and G6PD Deficiency with Particular Reference to the Malaria Hypothesis. B ulletin of the World Health Organization 34: 379-93. Siniscalco, M.; L. Bernini; B. Latte; and A. G. Motulsky 1961 Favism and Thalassemia in Sardinia and Their Relationship to Malaria. Nature 190: 1179-80. Spielberg, S. P.; L. A. Boxer; L. M. Corash, j. D. Schulman 1979 Improved Erythrocyte Survival with High Dose Vitamin E Therapy in Chronic Hemolyzing G 6 P D and Glutathione Synthetase Deficiencies. Annals of Internal Medicine 90:53-54.


Nutritional and Biopsychological Constraints
THE ESSAYS OF PART III DEAL WITH GENETICALLY DETERmined biopsychological factors that affect foodways from the standpoint of nutritional adequacy and"taste." From a materialist perspective, it is axiomatic that human foodways that lead to severe malnutrition will tend to be selected against during both biological and cultural evolution. But this leaves us with the question of how much of what kinds of nutrients in what proportions and at what stage of life for males and females (including pregnancy and lactation) with different somatotypes and under varying conditions of health will result in a severely malnourished population. As the essays by Pellet and Lieberman make clear, nutrition is far from being an exact science. Not only are recommended daily allowances extrapolated from experiments on non-human models and from causally ambiguous epidemiological data, but to a considerable extent they express the outcome of political infighting between representatives of developed and developing nations as well as that between the vested interests of the agri-businesses and medical-nutritional establishments. Attempts to relate foodways to minimum nutritional requirements are further complicated by the existence of some innate taste preferences and avoidances. It is possible that these biopsychological features account for the overconsumption of certain foods (e.g., sugar) or the selection of certain less nutritious foods when more nutritious ones are available (e.g., candy versus grains). Yet, as Rozin's essay shows, cultural conditions can readily turn innately aversive substances l£ke hot chili peppers into the dominant taste signature of whole cuisines. The interplay between cultural conditioning and nutritional requirements is especially baffling in relation to the widespread preference for animal food, whose existence is documented in Abram's essay. As Lieberman's essay shows, from a strictly nutritional point of view, neither animal nor plant foods can be said to be more suitable as a source of essential nutrients, nor can either be said to be free of


III. Nutritional and Biopsychological Constraints

nutritional risks and penalties when overconsumed. But it should be kept in mind that the evolution offoodways is probably subject to selection by costs and benefits that encompass more than the strictly nutritional composition offoods. Thus, the preferences for animal foods, especially for meat, may be a response to the higher density" of calories and nutrients packed into a gram of meat versus a gram of plant food. We will return to this issue in Part IV.



Problems and Pitfalls in the Assessment of Human Nutritional Status
MORE THAN 40 SEPARATE ELEMENTS OR COMPOUNDS HAVE NOW been identified as necessary for life, together with at least 40 grams per day of amino acids as protein and more than 200 grams per day of a mixture of carbohydrate, protein, and fat that can be metabolized for energy. Although all these materials are required for daily metabolism to proceed, it is not necessary that they all be provided daily in the diet. The ability of the body to draw on reserves or stores allows varying amounts of time to elapse before deficiencies may be recognized. Oxygen (if that be a nutrient) is needed continuously, water deprivation cannot last for more than a few days; total food deprivation can last for weeks to months, depending on fat reserves; whereas vitamin A stores may last for more than a year on essentially zero intake (Hume and Krebs 1949). Daily requirements for nutrients also vary enormously, ranging from microgram quantities for Vitamin B12 , milligram quantities for many minerals and most vitamins, gram quantities for individual essential amino acids, calcium and phosphorus, and, of course, several-hundred-gram quantities of energyproviding materials. The distinction between required and non-required nutrients is not always clear-cut. Fiber is necessary in the diet for optimal health, yet it is not a nutrient; its function, indeed, is not to be absorbed. Several mineral elementsfor example, zinc, copper, and selenium-have only relatively recently been demonstrated as essential nutrients, but nickel, tin, and vanadium remain in the uncertain category. Of the essential amino acids, histidine is now known to be required by adults, but the demonstration of deficiencies can take months, since the amino acid can be synthesized in the body, but only at a rate somewhat below requirements. Normal individuals can manufacture as much cystine and tyrosine, as they need; thus, these two amino acids are termed non-essential (non-essential amino acids are so essential that the body has learned to synthesize them!). Under conditions of liver damage, however, cystine and


III. Nutritional and Biopsychological Constraints

tyrosine can no longer be synthesized fast enough from methionine and phenylalanine respectively, so that they become essential and must be supplied in the diet (Horowitz et al. 1981). With some vitamins, synthesis may occur (invalidating some of our definitions): the well-known example of nicotinic acid, synthesized from the amino acid tryptophan, complicated the search for the cause of pellagra (Horwitt et al. 1955). Vitamin D (cholecalciferol) can be synthesized in the skin, and it functions more as a hormone than as a vitamin (DeLuca 1974). Even vitamin C, the first discovered if not first named of the vitamins, whose lack has probably influenced human history more than any other deficiency, may be synthesized to a very limited degree in pregnancy and lactation (Rajalakshmi et al. 1965). Diets (and most individual foods) are mixtures of nutrients, in varying proportions, along with other compounds such as fiber (of various types), phytates, cholesterol, natural toxicants, and trans-fatty acids, which may be harmful or beneficial. No single food can supply all our needs for long, and the lack of intake of a single food mayor may not affect the intake of nutrients. A safe rule of thumb is that the more components there are in a dietary, the greater the probability of balanced intake. The corollary of this is that monotonous dietaries with few components are more likely not only to show deficiencies, but to be seriously affected by dietary avoidances. The assignment of credit or blame to dietary intake for the development or prevention of malnutrition is much less straightforward than many assume, both inside and outside the field of nutrition. Nutritional status and health status overlap; adaptation to high or low intakes may occur; nutritional requirements, although based on scientific facts, depend on informed judgments and are subject to a wide range of individual variability; and, finally, estimates of nutrient intakes are only approximate except under controlled metabolic conditions that are hardly nonnal. It is thus not surprising that the role of food preferences and food aversions in nutrition must often remain anecdotal and equivocal. It is now increasingly recognized that malnutrition may be caused less by nutrient deficiency as such and more by many interrelated social, political, and economic factors Oohnsson 1981; Pellett 1983); the widespread prevalence of malnutrition is usually a symptom of a very sick society (Maletnlema 1980). Because of this multifactorial causation, solutions must also be multifaceted, even if such measures as the elimination of poverty and improvement of living standards are basic. Malnutrition affects the growth, development, and survival of children and the health, activity, and well-being of adults. Conventional solutions in the form of specific programs are usually inadequate, and their effects are transitory, since they do not reach causes. Nutrition can be improved through upgrading living standards-particularly the level of real income, food availability, and health services. Solutions will be found, therefore,


6. Problems and Pitfalls in the Assessment of Human Nutritional Status

mostly outside the traditional nutritional field, in economic and social development. Although the tenn "malnutrition" strictly should include overnutrition and some of the diseases of affluence, it is often used only to refer to the condition resulting from a deficient intake of energy or of a particular nutrient. Four especially important and broad causes of malnutrition are the following: 1. Insufficient supply of the foods necessary for a balanced diet, often due to production failure, poor soil, climate, and fanning techniques, or overpopulation 2. Uneven distribution of the available food (both between and within families) 3. Lack of knowledge about nutrition and health 4. Poor health and sanitation, including infectious diseases that are synergistic to malnutrition (Latham 1979) It must, of course, be recognized that these causes are not separate and distinct causes, but overlap with each other. The characteristics of the major world nutritional disorders are summarized in Table 6.1. The fundamental roles of health, wealth, and sanitation factors must be recognized. Nevertheless, depending on the disorder, actions can be taken to improve the nutritional status of the most vulnerable groups, which are women of child-bearing age and young children. The causes of malnutrition for individuals or communities can be grouped into the levels summarized in Figure 6.1. These levels are related to moneybased societies, but similar levels, overlapping in various ways, exist in other societies. Under present-day economic conditions, causes of malnutrition start with overall food availability. This is affected by international and national political-economic activities as well as agricultural policy both within and external to the country. The next determinant is family purchasing power. This also is dependent on political and economic factors, but at a more local level. Thus, what food is purchased is determined both by food availability and by economic status and money availability. At the next level comes food choice. Within any economic group, the pattern of food purchases is dependent on likes and dislikes as well as the relative prices of the various foods. Nutrition education can work at this level by encouraging selection of nutrient-rich foods over nutrientpoor sources with the same cost. However, nutritional education has been far less effective than we would wish (Schiirch 1983). The next level below the family food-purchasing pattern, and overlapping it to some degree, is the pattern of food distribution within the family. This is dependent on culture and is frequently a major cause of malnutrition in the vulnerable groups. Women's and children's greater need for protein and micro-


TABLE 6.1. The Characteristics of Hunger and the Major Nutritional Disorders
Protein-Energy Malnutrition (PEMb) Nutritional Marasmus

Causation/ precipitation Long-term

Hunger a


Poverty, poor agriculture

Low-protein diet


Poverty, crop failure, war

Early weaning, infections Children under 1 year


Vulnerable groups and main age of incidence Major features

All ages

Children between 1 and 2 years

Growth failure, wasting, lethargy


Edema, fatty liver, reduced serum albumin


Reduced growth, reduced work capacity, high mortality

High mortality, impaired mental development

High mortality, impaired mental development

Areas of incidence

All areas of poverty

Urban areas, large families with low incomes and poor education

Rural Africa, areas with low-protein weaning foods



Widespread in poor areas

Increasingly rare in Middle East and North Africa




Iron-deficiency Anemia

Low Birth-Weight (LBW) Infants c

Low intakes of carotene and/or retinol Early weaning, infections Preschool children

Low intakes of iodine

Low intake/ absorption of iron

Poor dietary intake since conception, infections of mother

Blood loss from infestations Older children, females Children « 3 years) and females of childbearing age

Low weight gain in pregnancy Mothers of poor socioeconomic status

Night blindness, xerosis of conjunctiva and cornea, keratomalacia, low serum retinol High mortality especially when associated with PEM, blindness Rice staple areas, areas of poverty

Enlarged thyroid

Low hemoglobin (microcytic hypochromic anemia if severe)

Hypoglycemia, hypothermia, poor resistance to infection (low IgG)


Pallor, reduced work and learning efficiency

High mortality, suboptimal development, high incidence of infection

Areas with low soil iodine, mountain areas and certain oases


Low socioeconomic areas with high prevalence of infections

Not widespread in wheat-staple areas of Middle East and North Africa

High in localized areas


Incidence highly correlated with socioeconomic indicators-possibly 20 percent of all births in poor areas


III. Nutritional and Biopsychological Constraints

TABLE 6.1. (cont.)
Protein-Energy Malnutrition (PEMb)
C haracteristics
Measures to eliminate

Hunger a
Major socioeconomic, agricultural, and educational improvements

Nutritional Marasmus
Major socioeconomic and educational improvements, control of infections

Control of infections, higher protein in weaning foods

SOURCE: Pellett (1983). aThere is a considerable degree of overlap of causation, consequences, and prevalence of proteinenergy malnutrition with low birth weight infants. b PEM when early or of mild to moderate severity is usually sub-clinical and can only be diagnosed by anthropometric criteria (weight/age, height/age, and weight/height). c Infants of birth weight below 2,500 grams. In developing countries the majority of cases are due to fetal growth retardation.

nutrients, despite their lower food-energy needs-that is, their need for a higher concentration of nutrients-is often difficult to impart, since nutritional education will frequently conflict with cultural nonns. Finally, because of the now well-known inter-relationship between malnutrition and infection (Chen, 1983; Scrimshaw, Taylor, and Gordon 1968), consumed food may not be fully utilized. An individual suffering from infection (or infestation) may not only have a reduced food intake, but may also have poorer utilization for a range of nutrients. Sanitation and the availability of clean water thus profoundly affect nutritional status. Techniques for evaluating the nutritional status of populations can also come from a consideration of various levels of infonnation. These can be separated into two major categories, the first pertaining to agriculture and food availability, the second pertaining to the health of the population. These are shown in Table 6.2. Agriculture and food production data have limitations, but they can indicate the approximate availability of food supplies and nutrients to a population. Close examination of agricultural production data can also allow judgments on the success or failure of agricultural techniques. The next level of infonnation concerning food derives from dietary surveys and food consumption patterns within the society and provides data about socioeconomic variables and


6. Problems and Pitfalls in the Assessment of Human Nutritional Status

Increased consumption of green vegetables, fortification, improved socioeconomic conditions

Iodine enrichment

Iron-deficiency Anemia
Iron enrichment, control of infestations

Low B irth-Weight (LBW) Infants e
Maternal nutrition programs, prevention of infections, major socioeconomic and educational improvements

FIGURE 6.1. Major Factors Affecting Nutritional Status
Food availability

International and national politics and economics, agricultural policy, production and distribution Political and economic factors at a local level Targeted economic assistance

Family IurChaSing power


Improved purchasing power

Family food-purchasing pattern

Poor nutrition knowledge


Nutrition education

Improved food selection

Witrun-rmilY food distribution

Poor nutrition knowledge Nutrition education, targeted food assistance

Improved food distribution

Utilization of foods by consumer

Infection, infestation, poor sanitation Health advice and services ~ Improved food utilization

Individual nutritional status SOURCE: Sequence adapted from Pinstrup-Anderson (1982).



III. Nutritional and Biopsychological Constraints

TABLE 6.2. Infonnation Needed for the Assessment of Nutritional Status
Food-Related Data Category

Health-Related Data Country
Vital and health statistics

Food balance sheets

Dietary surveys

Anthropometric, biochemical, clinical testing


Approximate availability of food supplies per capita

Approximate nutrient intakes of individuals and comparison with nutritional requirements

Morbidity and mortality data and degree of risk to the community

Effect of nutrition on physical development, biochemical function, and development of clinical abnormalities

the distribution and storage of foods. Dietary surveys, though difficult to perform with accuracy, can give the most detailed information on food and nutrient consumption at the family level. Conversion of food intake to nutrient intake is now performed rapidly through the use of computers, but is still dependent on both judgment and the accuracy of food-table data. Once nutrient intake data are established, these must be compared with standards before decisions can be made concerning the nutritional value of the diet. Derivation of the appropriate requirement standards and comparison with intake values does not always allow clear-cut conclusions concerning either the nutrient or the diet as a whole. A recent definition of nutritional requirements comes from the Canadian recommended nutrient intakes (Bureau of Nutritional Sciences 1983). "Requirement" is taken to mean that level of dietary intake of a nutrient that meets an individual's need, defined as the establishment and maintenance of a reasonable level in tissues or stores in the body. In the case of energy, the requirement is considered to be the level of intake needed to maintain appropriate body size and composition and expected work and leisure activity. For infants, children, and pregnant women, the requirement is meant to include needs for the deposition of new tissues, and for lactating women, the production of milk. A similar definition is used for the recommended dietary allowances (RDAs) in the


6. Problems and Pitfalls in the Assessment of Human Nutritional Status

United States, which appear at four- to five-year intervals. The most recent of these appeared in 1980 (NAS-NRC 1980) although a 1985 version has been prepared and is awaiting publication. All of these definitions refer to the nutrients needed to maintain health in already healthy individuals. Among normal individuals, requirements for many nutrients are affected by the nature of the diet, body size, activity, age, sex, and physiological state. There is thus individual variability, and the requirements for a group will show a range, although in tabular presentations an average for a set of reference individuals is often shown. RDA is supposed to take into account individual variability and to meet the requirements of almost all individuals in a group of specified characteristics. For food energy, in contrast, average requirements are specified because both inadequate and excess intakes of energy sources may present a risk to health. Because of this major difference, food energy was removed from the main summary table of the 1980 edition of the RDAs, and average food energy allowances are now tabulated in the energy section, where they are shown together with a range for adults of ±400 kilocalories per day. Thus, the approach to drawing conclusions from the average intakes of groups of individuals will differ for energy and for other nutrients. Human protein and energy needs are addressed in depth about every decade by an international group under the auspices of the Food and Agricultural Organization (FAO) and the World Health Organization (WHO). The United Nations University (UNU) collaborated in the most recent report (FAO /WHO/UNU 1985). Although the actual recommended allowances for protein and energy do not differ greatly from previous recommendations, the basis used for both recommendations has changed. The new report notes the apparent paradox of the survival of populations that consume only 80 percent of their estimated energy needs. These populations adapt to these lower energy intakes by a reduction in activity. This has been documented by Viteri (1976) for Guatemalan plantation workers, whose productive capacity is limited by their food intake but who remain almost inactive when their day's work is completed. Mter additional food was provided, both consumption and activity increased. The new report emphasizes the importance of adequate food for discretionary as well as work activities and underlines the obvious (but often ignored) fact that it is impossible to establish an appropriate average figure for a population's energy requirements unless their activity level is also specified. Not only do individuals adapt, but whole societies may develop patterns compatible with low food intakes, which may not, however, be compatible with their long-term interests. The report suggests a simplified approach to estimating energy needs in which needs are defined as follows: food energy need == basal needs x activity,


III. Nutritional and Biopsychological Constraints

where basal needs are calculated from a fonnula that includes sex, age, and weight and an appropriate activity factor is either calculated from daily activity patterns or assigned. These assigned factors would range from 1.4 for light activity to 2.2 for heavy activity. It has been recognized for some years that the daily allowance of 0.57 grams per kilogram of high-quality protein (egg, fish, milk) recommended by the FAO/WHO (1973) committee was too low (Garza, Scrimshaw, and Young 1977). The FAO/WHO/UNU (1985) committee reviewed a number of short (in tenns of weeks) and long (in tenns of months) protein requirement experiments in humans and concluded that the new safe protein allowance should be o. 75 grams per kilogram, again expressed in tenns of eggs, fish, or milk. This is an apparent increase of 30 percent. In practice, recommended allowances will increase far less, because there had previously been an excessively large correction factor for protein quality, net protein utilization (NPU), a measure that takes into account digestibility and amino acid composition (see Chapter 9). If the N P U for mixed ordinary diets is assumed to be 70 percent, then the allowance of 0.5 grams per kilogram for high quality protein is increased to 0.8 grams for mixed dietary protein. Newer values for adult human amino acid requirements would now increase the N P U of the same mixed dietary protein to 90 percent, and thus the new allowance becomes 0.83 grams per kilogram, almost identical to the previous value. An interesting outcome for a rather divisive dispute-protein requirements have been increased, yet remain the same. The overall discussions have not significantly changed the now widely accepted view that many population groups can obtain adequate protein if they are able to obtain sufficient food; thus, agricultural policy should remain centered on total yields, and increases in protein concentrations in plant-breeding programs are not required (Nygaad and Pellett 1986). Despite this it remains true that there are many population sub-groups, especially among the poor, where protein intake may indeed be limited and increased protein is desirable. Direct comparison of dietary intakes with other standards, such as the RDAs, may still, however, lead to inaccurate conclusions. The problems categorized by the 1980-85 Committee on Dietary Allowances (N AS-NRC 1986), reached the following conclusions. (1) RDAs are for reference individuals and must be adjusted for individuals with other physical characteristics. (2) The RDAs are guidelines for separate essential nutrients, whereas people eat food, and rarely for the express purpose of consuming nutrients. No set of nutrient guidelines can adequately address these nonnutrient factors. (3) The guidelines are often incorrectly interpreted. RDAs cannot be used to determine the adequacy of the average consumption of a group. If a group's average consumption meets the RDAs, it cannot be concluded that all members of the group are well nourished, because the distribution within the group may be


6. Problems and Pitfalls in the Assessment of Human Nutritional Status

skewed; that is, some may be consuming more and some less. However, if the average intake is much below the RDAs, then it can be concluded that there may be a significant risk of malnutrition. Failure to meet the RDA in anyone day should not be interpreted as indicating poor nutrition, since the RDAs are set for intakes over time. Individual intakes below the RDA also cannot be interpreted as indicating poor nutritional status, since RDAs are set above the needs of many individuals. It is often assumed that the RDAs (except for energy) are set to meet the needs of 97.5 percent of the population. This is strictly true only for protein, where the coefficient of variation from a large number of requirement studies is known. For other nutrients the "safety factor" used is not consistent and differs in derivation and magnitude from nutrient to nutrient. A policy analogous to "caveat emptor' probably applies to all users of all RDAs. In addition to simple extraction of allowances from the summary table, careful reading of the text is required. The question whether the dietary survey method followed is appropriate may also arise. Vitamin AI carotene, for example, is present in high levels in relatively few foods and may thus be consumed only infrequently. For such nutrients, ordinary dietary recall methods, even when three-day recalls are used, may be misleading. Some vitamins are stored in the liver, so that nutritional status can be maintained if high levels of intake occur perhaps once or twice per month. The metallic micro-nutrient copper may also be in this category. For circumstances such as these, food frequency determination may permit more accurate predictions of nutritional status than dietary surveys. Vitamin A status can also be used as an example of how overall dietary circumstances may affect the relative importance of dietary preferences. Many children throughout the world have an aversion to green leafy vegetables. This aversion to the only cheap source of vitamin A activity (in the form of several carotenes of varying potency) may be life-threatening to children in Asian riceeating communities and can also lead to high levels of xerophthalmia and blindness. For a child in the United States or Europe whose vitamin A is readily obtained from animal foods, the same aversion is of little consequence. The paradox of blindness and vitamin A deficiency in individuals surrounded by lush, green vegetation rich in pro-vitamin A activity has been noted by McLaren (1963). Observations on the health of populations are also used to indicate nutritional status, although impairment of health can usually be directly ascribed to faulty nutrition only when corresponding food and nutrient intake data are available to correlate with the health data. These data start at the regional or country level, including vital and health statistics when they are available. These can identify the extent of risk to the community. At a more individual level, nutritional assessment includes anthropometric, biochemical, and clinical studies. Consid-


III. Nutritional and Biopsychological Constraints

eration of the results of such surveys can give information on the effects of nutrition on physical development, on the impairment of biochemical function, and on the deviation from health due to malnutrition. Assessment of nutritional status by newer functional tests (Solomons and Allen 1983) is an important innovation that may allow more specific conclusions to be drawn concerning nutrient deficiency and health. Poor apparent nutritional status-for example, as measured by inadequate growth and development of children-may be related more to infective disease or inadequate water supply and sanitation than directly to food availability and food selection. Analysis is further complicated by the synergistic relationship among these factors. Even the type of measurement used for anthropometry will often allow varying conclusions to be drawn (Table 6.3). Weight-for-age measurements in children may overestimate the prevalence of malnutrition when compared with weight-for-height measurements. Children from poor environments may weigh less than well-nourished children of the same age but, having adapted to the situation, may be apparently healthy. Are these children suffering from malnutrition or not? The answer depends on the definition used. Smaller overall size but balanced weight for height may be a reflection of adaptation to a lower level of total food energy intake. Adaptation can also occur for other nutrients. Two populations with large differences in protein intake may both be in nitrogen balance and meeting their requirements. Similarly, the population of New Zealand attains selenium balance at an intake much below that of the United States with no apparent effects on health (Levander 1982). For iron, the situation is even more complex. The RDA for iron is given as 10 milligrams per day for an adult male and 18 milligrams for an adult female. This, however, tells very little about how much iron is required in diets, because both the source of iron and the presence of phytates, oxalates, tannins, and crude fiber will greatly influence the biological availability of dietary iron. In general, iron of vegetable origin is poorly available, ranging from as low as 1 to 5 percent for leaf sources such as spinach, lettuce, and cassava leaves. In contrast, the heme iron in red meat is at least 10 to 20 percent absorbed and also improves the absorption of vegetable iron in the diet (Monson et al. 1978). Iron absorption is enhanced by ascorbic acid in the diet and reduced by protein deficiency. The adequacy of dietary iron intake is complicated still further by environmental factors. Hookworm, schistosomiasis, and malaria can increase iron loss from the body, and enteric infections may further decrease iron absorption. Clearly one can tell little about the adequacy of dietary iron intakes per se or the probability that they will be associated with the known consequences of iron deficiency, which can include reduced physical capacity and work performance and impaired resistance to infection, resulting in increased sickness.


6. Problems and Pitfalls in the Assessment of Human Nutritional Status

TABLE 6.3. National Nutrition Survey of Egypt: Anthropometric Criteria Used, in Percent
Normal Moderately malnourished Severely malnourished

Weight for Height
95 2 <1

Height for Age
78 17 5

Weight for Age
53 38

SOURCE: AID (1978). NOTE: The subjects of the survey were 8,016 children (4,240 males and 3,776 females) below the age of six. "Normal" was defined as above 85 percent of the reference standard population for weight/height and above 90 percent for height/age and weight/age. "Moderately malnourished" was between 80 and 84.9 percent of the standard for weight/height, 85 to 89.9 percent for height/age, and 75 to 89.9 percent for weight/age. "Severely malnourished" was less than 80 percent of the standard for weight/height, less than 85 percent for height/age, and less than 75 percent for weight/age. The last category thus includes Gomez's second and third degree malnutntion (Gomez et a1. 1956).

Errors in the assessment of vitamin A nutriture from food intake values can readily occur, since the values given for the vitamin A content of vegetable foods in some food tables may be misleading. The biological requirement for vitamin A can be met either from pre-fonned vitamin A (retinol), present in animal foods, or from various carotenoids, such as ex carotene, r3 carotene, and cryptoxanthin. The degree of conversion to retinol for various carotenes differs markedly and ranges from zero to about 17 percent (r3 carotene). Of over 500 carotenoids found naturally, only about 50 show pro-vitamin activity, and of these r3 carotene is the most important quantitatively in human nutrition (Olson 1983). From this it follows that color bears no direct relation to vitamin A activity: for example, a pale yellow com may supply more pro-vitamin A than some more deeply colored varieties, and although red palm oil is a good source of r3 carotene, the majority of the pigments present are non-active. Newer food table values attempt to allow for the varying activities of vegetable carotene sources by giving retinol equivalents for vitamin A activity: different factors are used to convert each of the carotenes present to retinol. Yet problems arise here too, since few complete analyses exist showing the distribution of the various carotenes in vegetable foods. Conversions are also complicated by the fact that the international unit (I U) for carotene is not the same as the I U for retinol. Guidelines for conversion are given in the 1980 RDA manual (N ASNRC 1980). Knowledge that vitamin A may be toxic at excessive levels may also lead to problems. Although it does not follow that vegetables with high levels of carotene pigments (e. g., sweet potatoes) are also toxic, the high


III. Nutritional and Biopsychological Constraints

vitamin A scores given such foods in some food tables may lead to that impression. In addition, values, especially for minerals and vitamins, differ from region to region within a foodstuff (e. g., outer and inner leaves), and marked seasonal variations in content may also occur. Thus, complexities exist even in apparently simple food tables, and judgment and experience are needed in the interpretation of these data. Certainly food habits, including preferences and aversions, may affect nutritional status, but positive or negative relationships may be extremely difficult to prove except in extreme instances. Among the factors that may make conclusions difficult are the following: 1. The multifactorial etiology of malnutrition 2. The ability to adapt to higher or lower than optimal intakes, and the varying definitions of optimal 3. The varying bioavailability of nutrients 4. The difficulty of assessing food intake with accuracy 5. Problems related to the accuracy, variability, and availability of nutrient composition data, especially for unusual foods 6. The relatively "soft" nature of requirement values for many nutrients The general tenor of this brief paper may seem to be negative and to lead to the conclusion that everything is too complicated and that no relationships between food preferences and nutritional status can be drawn. General views on the causation of malnutrition have undergone considerable change over the last two decades, and vast numbers of inter-relationships and inter-dependences have been recognized. The science of nutrition, in common with many other disciplines, has undertaken much soul-searching with regard to its place and ,Ele in the world. Perhaps the expectations were too high, but the difficulty of action in the real, highly complex world is now recognized. Problems of malnutrition are so closely related to the major problem of society that it is easy to be overwhelmed by the size of the problem and to feel that one is powerless to make any impact. The pessimistic, perhaps realistic, view is that the elimination of malnutrition will be accomplished only when poverty and injustice are overcome and health care is available for all. This is likely to be far in the future, and problems of malnutrition persist and are increasing. Suggestions for nutritional improvement are thus now returning from vast planning endeavors to smaller targeted-assistance programs aimed at those at most risk in society, such as the mother and the young child. To paraphrase a biblical saying, man may live by bread alone, but his wife and children cannot. Here, an understanding of the role of food preferences and aversions in the selection of an adequate diet for the mother may be of supreme importance. This can not only decrease the prevalence of low-birth-weight infants but also help with the continuance of


6. Problems and Pitfalls in the Assessment of Human Nutritional Status

breast-feeding and thus positively influence infant nutrition and reduce infant mortality from two directions simultaneously.

References Cited
Agency for International Development (A I D) 1978 Arab Republic of Egypt National Nutrition Survey. Washington, D. C.: U.S. AID, Department of State. Bureau of Nutritional Sciences 1983 Recommended Nutrient Intakes for Canadians. Ottawa: Canadian Government Publishing Centre. Chen, L. 1983 Planning for the Control of the Diarrhea-Malnutrition Complex. In Nutrition in the Community, 2d ed., D. S. McLaren, ed., pp. 134-60. Chichester and New York: Wiley. DeLuca, H. F. 1974 Vitamin D: The Vitamin and the Hormone. Federation Proceedings 33:2211-19. Food and Agriculture Organization/World Health Organization (F A 0 / WHO) 1973 Energy and Protein Requirements. Report of a Joint FAO/WHO Ad Hoc Expert Committee. World Health Organization Technical Report Series no. 522. Geneva: WHO. Food and Agriculture Organization/World Health Organization/United Nations University (FAO/WHO/UNU) 1985 Energy and Protein Requirements. Report of a]oint F AO/WH O/UNU Expert Consultation. Geneva: WHO. Garza, D.; N. S. Scrimshaw; and V. R. Young 1977 Human Protein Requirements: A Long-Term Metabolic Nitrogen Balance Study in Young Men to Evaluate the 1973 F AO /WH 0 Safe Level of Egg Protein Intake. Journal ofNutrition 107:335-52. Gomez, F., et a1. 1956 Mortality in Second and Third Degree Malnutrition. Journal of Tropical Pediatrics 2: 77-83. Horowitz, J. M.; E. B. Rypins; ]. M. Henderson; S. B. Heymsfield; S. D. Moffitt; R. P. Bain; R. K. Chawla; J. C. Bleier; and D. Rudman 1981 Evidence for Impairment of Trans-Sulfuration Pathway in Cirrhosis. Gastroenterology 81:668-75. Horwitt, M. K. 1955 Niacin-tryptophan Relationships in the Development of Pellagra. American Journal of Clinical Nutrition 3:244-45. Hume, E. M., and H. A. Krebs (comps.) 1949 Vitamin A Requirement of Human Adults. Report of the Vitamin A Subcommittee of the Accessory Food Factors Committee. Medical Re-


III. Nutritional and Biopsychological Constraints

search Council (Britain), Special Report Series no. 264. London: HMSO. ]ohnsson, U. 1981 The Causes of Hunger. (UNU /WHP) Food and Nutrition Bulletin 3(2): 1-9.

Latham, M. C. 1979 Human Nutrition in Tropical Africa. Food and Nutrition Series, no. 11, rev. 1. Rome: Food and Agricultural Organization. Levander, O. A. 1982 Selenium: Biochemical Actions, Interactions and Some Human Health Implications. In Clinical, Biochemical, and Nutritional Aspects of Trace Elements: Current Topics in Nutrition and Disease, vol. 6, A. S. Prasad, ed., pp. 345-68. New York: A. R. Liss. McLaren, D. S. 1983 Malnutrition and the Eye. New York: Academic Press. Maletnlema, T. N. 1980 How Europe Can Contribute to Nutrition Research for Developing Countries. In Nutrition in Europe: Proceedings of the Third European Nutrition Conference, Leif Hambraeus, ed., pp. 10-15. Stockholm: Almqvist and Wiksell International. Monsen, E. R. L.; L. Hallberg; M. Layrisse; D. M. Hegsted; ]. D. Cook; W. Mertz; and C. A. Finch 1978 Estimation of Available Dietary Iron. American Journal of Clinical Nutrition 31: 134-41. National Academy of Sciences-National Research Council (N AS-NRC) 1980 Recommended Dietary Allowances. 9th rev. ed. Washington, D. C.: NAS. Recommended Dietary Allowances. 10th rev. ed. Washington, D. C.: 1986 N AS. (Publication delayed.) Nygaad, D., and P. L. Pellett (eds.) 1986 Dry Area Agriculture, Food Science, and Human Nutrition. New York: Pergamon Press. Olson, ]. A. Formation and Function of Vitamin A. In Polyisoprenoid Synthesis, vol. 2, 1983 ]. W. Porter, ed., pp. 371-412. New York: Wiley. Pellett, P. L. 1983 Commentary: Changing Concepts on World Malnutrition. Ecology of Food and Nutrition 13: 115-25. Pinstup-Andersen, P. 1982 Introducing Nutritional Considerations Into Agricultural and Rural Development. Food and Nutrition Bulletin 4:33-41. Rajalakshmi, R.; A. D. Doedhar; and C. V. Ramakrishnan 1965 Vitamin C Secretion During Lactation. Acta Scandanavia 54:375-82. Schurch, B. (ed.) 1983 Evaluation of Nutrition Education in Third World Communities. A Nestle Foundation Workshop. Bern and Vienna: Hans Huber.


6. Problems and Pitfalls in the Assessment of Human Nutritional Status

Scrimshaw, N. S.; C. E. Taylor; and]. E. Gordon 1968 Interactions of Nutrition and Infection. World Health Organization Monograph Series no. 57. Geneva: WHO. Solomons, N. W., and L. M. Allen 1983 The Functional Assessment of Nutritional Status: Principles, Practice and Potential. Nutrition Reviews 41: 33-50. Viteri, F. E. 1976 Definition of the Nutrition Problem in the Labor Force. In Nutrition and Agricultural Development: Significance and Potential for the Tropics, N. S. Scrimshaw and M. Behar, eds., pp. 87-98. New York: Plenum Press.



Psychobiological Perspectives on Food Preferences and Avoidances

OUR WORK ON FOOD PREFERENCES AND AVOIDANCES COMES from a psychobiological perspective, but has been aided and enriched by information and perspectives from anthropology. I hope in this paper to show how a psychobiological approach can ask questions and provide information and points of view that can enrich the anthropological approach. I shall deal first with the biological roots of food choice in the human omnivore and then discuss the psychological dimensions of preference and avoidance. I will then describe what we know about how foods get to be liked or disliked and, in particular, how people come to like one innately aversive food, chili pepper. This is followed by discussion of a series of somewhat unrelated topics: disgust, the interaction of psychology, biology, and culture in determining food preferences, and the importance of traditional flavorings in understanding food choice, and then by some questions about the ways in which cultural institutions may be accounted for in terms of individual human psychology and biology. (See Barker 1982 for an excellent collection of papers representing the approaches presented here.)

The Mouth: Focus of Food Choice
Almost all material transactions for incorporating substances in the external world into the self involve the mouth, which serves as a guardian of the body (Rozin and Fallon 1981). It plays a pivotal role in accepting nutrients into the body and rejecting toxins. For most purposes, it is the last site at which reversible decisions about acceptance or rejection can be made. Once swallowing has


III. Nutritional and Biopsychological Constraints

occurred, the ingested substance is difficult to reject voluntarily. Given the special position of the mouth as the last reversible checkpoint and the great benefits and risks of putting food into the body, it should not be surprising that humans and other animals display a great deal of affect about things in the mouth. Our aim is to discover what gets into the mouth of any particular human, and why.

The Human Omnivore: The Biological Basis of Food Choice
Humans are quintessential omnivores. They are able and inclined to exploit a wide variety of food sources. Since it is not possible to specify in advance what sensory properties will characterize sources of nutrition (or toxins) in any particular environment, the omnivore must discover what is edible (P. Rozin 1976). That is, it is a fundamental part of omnivore biology to have only a few biologically based behavioral constraints or predispositions about foods. However, there are some genetically determined predispositions in all humans that promote adaptive food choice. The best-documented biological constraints are: an innate preference for sweet tastes and an avoidance of bitter tastes or irritation of the oropharyngeal surface (e. g., as produced by irritant spices or tobacco smoke) (Steiner 1979; see Cowart 1981 for a review); an interest in new foods, coupled with a fear of them; and some special abilities to learn about the delayed post-ingestional consequences of ingested foods (Booth 1982; P. Rozin 1976). The adaptive value of these biases is quite obvious, except perhaps the ambivalent attitude to new foods. This is simply a reflection of the risks and benefits of trying new foods: the great advantage of discovering a new source of nutrition, versus the danger of ingesting a toxinthe "omnivore's dilemma" (P. Rozin 1976; E. Rozin and P. Rozin 1981). Davis's (1928) classic work on the free selection of foods by three infants immediately after weaning confirms a strong preference for sweet items and an interest in variety in foods. These children did not show any evidence of a preference for meats. However, given the widespread preference for meat among adults, it is possible that there is a genetically programmed preference for meat that appears at some time after early childhood. Extensive research on food choice in rats, dating from the classic work by Curt Richter (1943), reveals the same biological biases we have discussed in humans as well as other biological biases that may be present in humans. Most notable is a genetically programmed enhanced liking for sodium salts under conditions of sodium deficiency (Beauchamp, Bertino, and Engelman 1983; Richter, 1956; P. Rozin 1976 for reviews).


7. Psychological Perspectives on Food Preferences and Avoidances

Use, Preference, and Liking
We must distinguish among three often confounded terms: use, preference, and liking (P. Rozin 1979). "Use" refers to whether and how much of a particular food a person or group consumes. "Preference" assumes a situation of choice and refers to which of two or more foods is chosen. One might prefer lobster to potatoes but eat more potatoes because of price or availability. "Liking," usually measured with verbal scales, refers to an affective response to foods and is one determinant of preference. Liking for a food almost invariably lneans that a person is pleased by its sensory properties (taste, smell, etc.). A dieter might prefer (choose) cottage cheese over ice cream, but like ice cream better. We tend to eat (use) what we prefer, and we tend to prefer what we like. However, availability, price, and convenience are critical determinants of use, but not of preference or liking. And the perceived health value of a food, while a potent determinant of preference and use (Krondl and Lau 1982), may have little to do with liking.

A Psychological Taxonomy of Preferences and Avoidances
Distinctions such as these require us to look systematically at what motivates people to accept or reject foods. By early adulthood, every human in every culture comes to adopt a culturally based set of beliefs and attitudes about objects in the world with respect to their edibility. We (Fallon and Rozin 1983; Rozin and Fallon 1980, 1981) have explored this psychological categorization of substances in American culture through interviews and questionnaire studies and concluded that there are three basic types of reasons (motives) for acceptance or rejection of potential food objects. Each of these reasons (Table 7.1) in one form motivates acceptance and in the opposite form motivates rejection. This simplified system emphasizes the principal feature motivating acceptance or rejection. We now briefly consider each of the three basic reasons.

Sensory Affective Factors
Some items are rejected or accepted primarily because of their sensory effects in the mouth (or sometimes their odor or appearance). We call such accepted items "good tastes" and those rejected, "distastes." Good tastes and distastes, by definition, produce appropriate positive and negative affect (like or dislike).


III. Nutritional and Biopsychological Constraints

TABLE 7.1. Psychological Categories of Acceptance (+) and Rejection (-)
Rejections Dimensions
Sensory-affective Anticipated consequences Ideational Contaminant Examples beer, chili, spinach allergy foods, carcinogens grass, sand feces, insects





SOURCE: Fallon and Rozin (1983). Reprinted with slight modifications, courtesy of Ecology of

Food and Nutrition.

Sensory-affective reactions can be innately attached to certain objects (acceptance of sweet tastes, rejection of bitter) or can be acquired. Substances that fall into the sensory-affective category for any individual are almost always acceptable foods in his or her culture. Individual differences on sensory-affective grounds (e.g., liking or disliking lima beans) probably account for most variations in food preferences within a culture.

Anticipated Consequences
Some substances are accepted or rejected as food primarily because of anticipated consequences of ingestion. These could be rapid effects, such as nausea or cramps, or the pleasant feeling of satiation. More delayed effects involve beliefs and attitudes about the health value of substances (such as vitamins or low-fat foods on the positive side, or potential carcinogens on the negative side). Anticipated consequences may be social, such as expected changes in social status as a consequence of eating a food. We place all acceptances based on anticipated consequences in the beneficial substances category, and all rejections for similar reasons in the danger category. There mayor may not be a liking for the taste of these items. If a liking (disliking) is present, it would be considered a secondary cause of acceptance (rejection).

Ideational Factors
Some substances are rejected or accepted primarily because of our knowledge of what they are and where they come from. Ideational factors predominate in


7. Psychological Perspectives on Food Preferences and Avoidances

Acceptances (;ood Taste
+ + +
saccharine medicines ritual foods



+ + +
Leavings of heroes or deities

many food rejections, but are less common in acceptances. Two clearly distinct subcategories of food rejection based primarily on ideational grounds can be distinguished. A large group of items can be classified as inappropriate. They are considered inedible within the culture and are refused simply on this basis. Grass and paper are examples. In such cases, there is not typically a presumption that these items taste bad, and the items are usually affectively neutral. They are inoffensive. In contrast, the category of disgust carries a strong negative affective loading. The nature or origin of the substance makes it offensive, nauseating, and contaminating (tiny amounts in an acceptable food render that food unacceptable). The sensory properties, including the taste, are disliked, even though in most instances the substance has never been tasted. Feces are a universal disgust. Within American culture, the category includes insects and dogmeat. We believe that this taxonomy has some cross-cultural validity in the sense that it captures the major motivations for accepting or rejecting foods. Of course, different items would fall into the various categories in different cultures. We explicitly recognize three limitations. First, some motivations for acceptance or rejection may have been overlooked. In particular, in some cultures a food is rejected on ideational grounds because of respect for it, as in the avoidance of some animals as food in parts of India (see, e.g., Eichinger Ferro-Luzzi 1975). Second, the taxonomy is not a classification of foods, but rather of motivations. The same food may fall into different psychological categories for the same person in different contexts, as when a charcoal-broiled hamburger moves from a distaste at breakfast to a good taste at dinner. Third, many foods are accepted or rejected for a combination of reasons and hence are not pure illustrations of the categories we have carved out.


III. Nutritional and Biopsychological Constraints

The Multiplicity of Motives in Acceptance and Rejection
The taxonomy lays out an array of motives that must be acknowledged if one wishes to understand and manipulate preferences. A food rejected because it causes illness has a very different psychological status from one that is rejected because it is offensive or tastes bad. Procedures to promote acceptance of a food would depend on the nature of the rejection. The set of anticipated consequences of ingestion alone contains many sub-motivations (a variety of negative or positive physiological or social effects). But the motivational frame for accepting or rejecting any potential food is yet more complex. Let us consider the case of coffee (Cines and Rozin 1982; Goldstein and Kaizer 1969). The motivations for consuming hot coffee are many in the United States, and include taste, a variety of positive caffeine effects, avoidance of caffeine withdrawal symptoms, social interaction (as in the coffee break), a desire to fill the stomach with a low-calorie substance, warming up, and so on. Most people in the United States consume a particular cup of coffee for a number of reasons, though for some there is a dominant reason. But even within the individual there is motivational variation. The first cup of the day, in the early morning, is often motivated primarily by its wakefulness-inducing properties, whereas later cups may have primarily social or other motivations. Ethnic groups in the United States differ in the motivational base for consuming coffee; for example, in the Philadelphia area, sensory (taste-smell) motivations are particularly important among Jews, whereas social factors seem more important among Italian Americans (Cines and Rozin 1982). Cross-culturally, there are differences in both specific motivations and the complexity of the motivation. In a Mexican highland village (Rozin and Cines 1981) weak but hot coffee is drunk once a day, an hour or so after awakening. The motivation for doing so is rather simple and uniform: to warm up in the morning. The point of this is that we must understand the motivational structure of consumption before attempting further analyses and general explanations. In terms of motivational structure, coffee is more complex than many other foods. The motivation for consuming rice, fish, or chili pepper is less variable, both among individuals within a culture and across cultures.

The Development of Likes and Dislikes
We will next consider the acquisition of likes and dislikes for foods. We will consider three contrasting pairs of categories (from Table 7.1). In each case, the first member of the pair involves an affective (taste like or dislike) response, and the second does not.


7. Psychological Perspectives on Food Preferences and Avoidances

Distaste Versus Danger
Aversion to bitter foods is innate. But many items that are initially acceptable come to be rejected, some because they come to taste bad, others because we learn that they are dangerous. We can understand that people avoid a food because they have been told that it will make them sick, but what makes them come to dislike its taste? When ingestion of a food is followed by nausea, humans develop a strong aversion to it (Garb and Stunkard 1974; Logue, Ophir, and Strauss 1981; Pelchat and Rozin 1982). This acquired bad taste can occur after a single negative experience with a delay of some hours between ingestion and its consequences. There is a strong tendency to develop a dislike for the taste or smell (as opposed to the appearance) of the food in question. There is also a tendency for a novel, as opposed to a familiar, taste to become the subject of such a learned taste aversion. Taste aversions will occur even if a person "knows" that the food he or she has eaten did not cause the illness. Nausea and vomiting are particularly potent in causing the acquired distaste (Pelchat and Rozin 1982). Other negative events such as hives, respiratory distress, headache, and cramps following eating usually induce avoidance motivated by danger rather than distaste. This contrast can be illustrated by the typical cases of two individuals who avoid shrimp. One has an allergy to shrimp and gets skin symptoms or respiratory distress after eating it. Such a person will avoid shrimp as dangerous but like its taste. If his allergy could be treated, he would be delighted to consume shrimp. The other person originally liked shrimp but got sick and vomited after eating it. This person will typically dislike the taste of shrimp and avoid it though realizing that it is in fact not dangerous, and was perhaps not even the cause of the sickness. There are limits to the explanatory power of the distaste-nausea linkage. It seems very unlikely that most acquired distastes have a history of association with nausea. Somewhat less than half of the people surveyed could remember even one instance of a food-nausea experience that led to an aversion. So we must presume that many people who dislike lima beans, fish, broccoli, and so on do so for reasons as yet undiscovered, and there are unfortunately no sound theories as to what the other paths to distaste might be.

Good Tastes Versus Beneficial Substances: The
Acquisition ofLikes for Foods
There is much more research and information about acquired likes than dislikes, but no single factor as potent as nausea for dislikes has been identified (see Beauchamp 1981; Booth 1982; P. Rozin 1983 for reviews). The overriding empirical relation in the study of acquired likes in all domains is that exposure


III. Nutritional and Biopsychological Constraints

tends to increase liking. Zajonc (1968) has suggested that exposure is a sufficient condition for liking-the mere "exposure" theory. Such effects have been demonstrated with foods (e.g., Pliner 1982), but there are limits; for example, overexposure can lead to dislike. Since cultural constraints normally arrange for the exposure of a child to available foods prepared in the traditional manner, mere exposure could account for a good part of the acquisition of liking for one's native cuisine. Alternatively, exposure could provide an opportunity for other processes to operate and produce liking. I will briefly review here the incomplete evidence we have on processes that might induce liking. The simplest explanation of increased liking with exposure is that exposure allows fear of a new food to dissipate; one learns that the food is safe (P. Rozin 1979). This does occur but by itself cannot account for a liking for (as opposed to lack of fear of) a food. The most straightforward parallel to the creation of distastes would be that some physiological consequence of ingestion, most appropriately an upper gastrointestinal effect (parallel to the nausea-distaste linkage), induces liking. Booth (1982; see also Booth, Mather, and Fuller 1982) has demonstrated a clear effect of rapid satiety: under some circumstances, hungry humans will come to prefer (in laboratory settings) a food with higher satiety value. Other positive effects of foods seem less potent in inducing liking, as indicated by the fact that people do not usually come to like distinctive-tasting medicines, even though they produce positive effects (Pliner et al. 1983). Pairing of a food with an already liked food can also lead to an enhanced liking for it. This Pavlovian-type process has been demonstrated once, for humans, in the laboratory. Flavors served with sugar (an already liked substance-a Pavlovian unconditioned stimulus) tend to become more liked (Zellner et al. 1983). There would certainly be cultural constraints to such a process: it is unlikely that a new meat could be made more acceptable for Americans by sweetening it with chocolate syrup. The production of enhanced liking through positive consequences or pairing with positive tastes is a process that should occur in animals as well as humans. In the animal literature on preferences, a consistent finding is that it is much harder to establish preferences than aversions (Rozin and Kalat 1971; Zahorik 1979). The difficulty in establishing food preferences in animals, in contrast to the many strong likes for foods in humans, suggests that some important factors operate only in humans. Sociocultural influences are the obvious candidate. We believe that social factors operate at two levels. First, social pressures (custom, the behavior of elders, the foods made available to the child) essentially force exposure, and as we have seen, exposure fosters liking. Second, the perception that a food is valued by respected others (e.g., parents) may itself be a mechanism for the establishment of liking. This has been directly


7. Psychological Perspectives on Food Preferences and Avoidances

demonstrated in a number of studies, the most convincing being experimental studies on pre-school children (Birch, Zimmerman, and Hind 1980). Preference for a snack is enhanced if it is given to the child in a positive social context, or used by the teacher as a reward. It is the teacher's indication that she values the food (using it as a reward) that seems to be critical. If the same food is given to other children, at the same frequency, in a non-social context, there is no enhancement of preference. The importance of the perception of social value in acquired liking is emphasized by the converse phenomenon. Research by Birch and her colleagues (Birch et a1. 1982) indicates that the perception by the child that others do not value the food per se and that they must be bribed to eat it (e.g., rewarded, told how healthy it is) seems to reduce the liking for the food. In other words, using the food as a reward indicates that it is valued and enhances liking. Rewarding the ingestion of the food is an indication that it is not valued for itself and tends to decrease liking for it. These results are compatible with self-perception theory (Bern 1967) in social psychology. The basic idea is that people "infer" their attitudes (e. g., liking a food) from their own behavior (e. g., choosing a food). If ingestion occurs without clear extrinsic motivation, there is a tendency to justify the behavior by increasing the value of the object. However, this relation would not hold if people interpret their behavior as externally constrained (e. g., they were forced to eat a food or there was nothing else to eat). The decline in value of an object when extrinsic reward is employed is called the "overjustification effect" (Lepper 1980). The notion that reward decreases or perhaps blocks liking is consistent with the fact that people are less likely to come to like oral medicines than foods, since medicine ingestion is clearly motivated by anticipated beneficial consequences (Pliner et al. 1983). It also "explains" the common claim that if something really tastes good, it is unlikely to be good for you.

Getting to Like Chili Pepper
Almost uniquely among animal species, humans regularly come to like substances that are innately aversive. Tobacco, coffee, the irritant spices, and the various forms of alcohol are among the more popular foods of humans around the world (P. Rozin 1978). The conversion of an aversion to a liking is one of the more striking phenomena in the area of human food selection. I will consider the processes at work in one of these instances, the acquired liking for chili pepper. Chili peppers produce oral pain and, at moderately high levels, induce defensive reflexes, including salivation, running of the nose, and tearing of the eyes. They have been consumed in Meso-America and other parts of the New World


III. Nutritional and Biopsychological Constraints

for thousands of years and were incorporated into many Old World cuisines following their discovery in the 16th century. The circumstances under which such an aversive food was rather readily adopted are mysterious, especially in light of the reluctance to adopt seemingly more nutritive and palatable products of the Americas, like corn and tomatoes. Chili peppers are among the most commonly consumed flavorings in the world and are probably eaten on a daily basis by over one-quarter of the adults in the world (P. Rozin 1978, Maga 1975 for general background). My emphasis will be on the mechanisms responsible for reversal of innate aversions. However, one can hardly refrain from discussing the possible adaptive value of consuming this popular spice. Chili peppers, per unit weight, are among the best sources of vitamins A and C in the world. Capsaicin, the substance that causes the mouth bum, activates the gastrointestinal system, stimulating salivation, gastric secretion, and gut motility. The role of any of these features in the adoption of chili pepper into new cultures or the acquisition of a liking for it is not known (see "Adaptation, Mechanism, and the Explanation of Culinary Practices" below). We have studied the liking for chili pepper in a Mexican highland village and in the University of Pennsylvania community (P. Rozin 1978; Rozin and Schiller 1980). At this time we do not know how this aversion reversal occurs, but I will summarize some of its basic features, and some constraints on theories of acquisition. A basic first point is that people consume chili pepper because they like it. It is a "good taste," and according to self-reports, is rarely consumed primarily because of its anticipated effects. Furthermore, people who like chili come to like the very same sensation (the mouth burn) that initially puts people off. In a Mexican village, children in the two- to six-year-old range receive gradually increasing amounts but are permitted to refuse it when it can be removed (e.g., by omitting the hot sauce on the tortillas). They are not rewarded in any obvious way for eating it; rather they observe that it is enjoyed by their elders. By age five to eight, most children in the village were voluntarily adding piquancy to their foods; they had come to like the "hot" stuff after months to years of exposure to it in a natural family setting. (American adults sometimes acquire likings for chili peppers very rapidly, after a few experiences.) Some of the mechanisms I have discussed may be at work, with initial exposure "forced" by mild social pressure to do what other members of the family do and the fact that moderate levels of chili are cooked into some of the foods. The mouth burn is paired with the repletion produced by the food in the rest of the meal, and with the already good tastes of the main food staples. Furthermore, the salivation facilitates mastication of an otherwise rather dry and mealy diet and may enhance the flavor of the food as well. The saliva and the pepper flavor and bum added to a rather bland diet seem to improve the taste of the food significantly, explaining why the most frequent explanation for chili eating offered by Mexicans is that it adds flavor or zest to food.


7. Psychological Perspectives on Food Preferences and Avoidances

Two possible explanations of the acquired liking for chili (and perhaps other innately unpalatable substances) depend on its initial unpalatability. The mouth pain of chili may become pleasant as people realize that it is not really hannful. This puts the pleasure of eating chili pepper in the category of thrill-seeking, in the same sense that the initial terror of a rollercoaster ride or parachute jumping is replaced by pleasure. People may come to enjoy the fact that their bodies are signaling danger while their minds know that there really is no danger (Rozin and Schiller 1980). Alternatively, the many painful mouth experiences produced by chili may cause the brain to attempt to modulate the pain by secreting endogenous opiates, morphine-like substances produced in the brain. There is evidence that, like morphine, these brain opiates do reduce pain. At high levels, they might produce pleasure. Hundreds of experiences of chilibased mouth pain may cause larger and larger brain opiate responses, resulting in a net pleasure response after many trials (P. Rozin 1982; Rozin, Ebert, and Schull 1982; see Solomon 1980 for a statement of opponent process theory that could account for this hypothesized effect). We are very far from an understanding of the development of liking for foods. My experience in trying to understand the liking for chili pepper convinces me that there are multiple routes to liking. We can find convincing arguments against every single mechanism we have been able to suggest. This indicates not that these mechanisms play no role in acquisition, but that there may be more than one way of producing an acquired liking. In the case of chili pepper and most other acquired likings, I am inclined to emphasize the importance of the social matrix and the valuation of the food by important others. A major role for social valuation might help to explain why non-human omnivores seem to develop acquired likings for food rather rarely, and why it is extremely difficult to get animals to like chili pepper. A few cases of chili-liking in animals have been reported; however, in all such cases, the animals (chimpanzees, dogs, macaque monkeys) were adored pets in homes where chili pepper was a frequent part of the diet. Perhaps these personal domesticates participated in the human social matrix that may be so important in the reversing of aversions and development of likings (Rozin and Kennel 1983).

Disgust Versus Inappropriateness
All of the categories we have considered up to this point probably exist in animals as well as humans. Disgust and inappropriateness-two ideationally based categories-require the mediation of culture. Inappropriate items (e. g., grass, paper, or sand) are those that are simply classified as non-foods by the culture. Their non-food status is learned by the child along with the acquisition of other types of infonnation about these items. In the case of disgust, infonnation provided through cultural mechanisms links up with a strong negative af-


III. Nutritional and Biopsychological Constraints

fective response. Objects falling into either category are rejected because of the idea of what they are and not necessarily because of their sensory properties or because they are dangerous. Disgusting objects have a number of other characteristics, in contrast to inappropriate objects. The critical difference between inappropriateness and disgust is that although inappropriate items are essentially neutral, disgusting items are offensive (Rozin and Fallon 1981). This offensiveness extends to the idea of the disgusting substance in the body, and to its sensory properties: taste, smell, and appearance. The thought of consuming disgusting substances-but not inappropriate ones-typically leads to nausea. Objects of disgust are contaminants. For example, most Americans would refuse a liked beverage if they believed that there was even a microscopic trace of urine in it. More generally, physical traces of disgusting substances render otherwise liked foods undesirable (trace contamination). Even pure ideational associations of a liked food with a disgust item, in the absence of any possible physical trace, can motivate rejection. For example, many Americans will not eat an otherwise liked soup if it is stirred by a brand new fly swatter (Fallon and Rozin 1983; Rozin, Fallon, and Mandell 1984). (Contamination seems to be essentially the same concept as Frazer's [1959] general principle of sympathetic magic, "contagion"- "once in contact, always in contact" [see Rozin and Fallon 1986 and Rozin, Millman, and Nemeroff 1986 for an elaboration of the relation between disgust and the laws of sympathetic magic].) There is also a distinct difference between the disgust and inappropriateness categories in terms of the items that fall into each. Objects of disgust are almost always of animal origin, and hence usually of substantial nutritive value (Angyal 1941; Rozin and Fallon 1980). Most tabooed foods, which we believe would usually have disgust properties, are animals or animal products (Angyal 1941; Tambiah 1969). Inappropriate items, by contrast, are usually vegetable or inorganic in nature. We (Fallon and Rozin 1983; Rozin and Fallon 1980) developed psychological profiles for the various psychological categories. Thus, disgust and inappropriateness induce different responses to items concerning nausea, contamination, negative sensory properties, and so on. Using this profile, we examined the psychological status of a number of foods. We found that pork for most kosher Jews, and meat for some vegetarians, had disgust properties. On the other hand, vegetables that some subjects describe as disgusting (e. g., brussels sprouts) actually have distaste, rather than disgust, properties (by our definition). Angyal (1941) suggests that disgust is fundamentally a fear of oral incorporation of an offensive substance. This is in keeping with the facial disgust response, which seems designed to keep offensive substances from the face, and particularly the area of the nose and mouth. In general, the more intimate the contact with a disgust substance, and hence the more real the threat of incorporation, the greater the disgust (Fallon and Rozin 1983).


7. Psychological Perspectives on Food Preferences and Avoidances

We do not have a satisfactory explanation of the origin of disgust. A major feature of such an explanation should be an economical description of the general nature of the objects of disgust (presumably more specific than identification of them as animals and animal products). I will briefly discuss here some conceptions of the basic nature of disgust objects and the mode of acquisition implied by each conception (see Rozin and Fallon 1986 for more details). One view is that at its core, disgust is an innate aversion to spoiled and decaying matter. From an adaptive point of view, this aversion could be related to the fact that decaying food is more likely to contain pathogenic organisms. However, we know of no evidence that non-humans (including other primates) systematically avoid decaying substances. Furthennore, there is evidence indicating an attraction to feces and to decay odors in young children (Senn and Solnit 1968; Stein, Ottenberg, and Roulet 1958), which disappears in the twoto seven-year age range. A second view focuses on feces as the primary disgust substance but, consistent with the development data, holds that the disgust is acquired (see, e. g., Angyal 1941 and the Freudian position generally). The universality of disgust for feces in adult humans argues in favor of this view (AngyaI1941). Given the initial attraction to feces, the emotionally laden toilet-training experience seems like the natural source for the origin of disgust. The conversion of an attraction into a strong aversion is a paradigmatic instance of what Freud described as a reaction fonnation (Senn and Solnit 1968). This position predicts the appearance of disgust around the time of toilet training and a generalization of disgust to other objects similar to feces on perceptual or conceptual grounds. A third view emphasizes filth or disorder as opposed to the narrower concept of decay or reference to a primal disgust substance. Association with filth (e.g., garbage, feces) would be a means for the extension of the category. Douglas (1966) explicitly identifies "filth" as disorder, anomaly, or matter out of place, taking a more conceptual-symbolic approach to this idea. A fourth view focuses on animalness (AngyaI1941). Angyal notes that as the animalness is removed from the offensive object, as by cooking or chopping, the object becomes less disgusting. Attempts have been made to describe what particular features of animals lead to disgust, with ideas such as anomaly (Douglas 1966), or closeness to humans in fonn (e. g., primates) or relation (e.g., domesticated pets) (Ortner 1973; Tambiah 1969). We have argued in accord with this view that all animals and their products have the potential to elicit disgust (Rozin and Fallon 1986). Finally, disgust can be related to the human social system, where it can be viewed as an emotionally laden aspect of human social relations. Many objects of disgust are human products or involve human mediation of one sort or another. Food is clearly a major means of social expression, a fact most clearly seen in the use of food in India (Appadurai 1981) or New Guinea (Meigs 1978, 1983). Incorporation of food produced, touched, or handled by a specific other


III. Nutritional and Biopsychological Constraints

can entail some notion of incorporation of this other. If so, the value of the food would depend on the nature of and the relation to that other (Appadurai 1981; Meigs 1978, 1983). With respect to hostile others, the food can have dangerous and sometimes disgusting properties. It is quite possible that all of the views described contain part of the truth. All, except the innate spoilage view, entail some sort of acquisition or enculturation process. It would seem most likely that verbal and non-verbal signals from parents and others-and perhaps negative facial expressions (disgust faces) in particular-playa role in communicating disgust to children. We are currently exploring the development of disgust and the idea of contamination in children (Fallon, Rozin, and Pliner 1984); surprisingly, we find that contamination does not appear until about 7 years of age in American children. We do not pretend that we have explained the mystery of how certain cognitions tap into the strong negative affect system that we call disgust. We know that when they do, a strong aversion results, which tends to be permanent and to resist modification by rational means. It is ironic that disgust presupposes a rather sophisticated cognitive base (in terms of categorization of the world and, perhaps, contamination) and yet is so resistant to further cognitive influence.

"Positive Disgust" and the Positive-Negative Balance
I noted above that, especially in animals, there is a strong bias to more rapid and more substantial learning about the negative properties of foods. An innate positive response to sweet tastes in rats can be changed into a strong aversion in one learning trial, but an innate aversion (as to bitter substances or irritants) is very difficult to reverse, and when there is success, the effect is small and the training extensive (e.g., Rozin, Gruss, and Berk, 1979; Zahorik, 1979). Presumably this bias is related to the destructive potency of many foods and other items in the environment and the adaptive importance of rapid learning about toxins, predators, and so on. The question then arises: how about humans? As I have pointed out, humans readily learn to reverse innate dislikes. Not only do humans come to like or even love many foods, but they develop strong likings for many types of objects (pets, people, sports teams, music, etc.). We seem in many ways to have overcome the negative bias. But have we? We stillieam aversions much more rapidly than preferences. Most critically, although there are a large number of items in the world that we reject as offensive (disgusting), and that can contaminate good food, there are practically no instances in American culture of the opposite of disgust-that is, substances eaten because of their nature or origin (as opposed to their sensory properties or the anticipated consequences of eating), which have the peculiar


7. Psychological Perspectives on Food Preferences and Avoidances

property that, in trace amounts, they can make a disliked food likable. As succinctly put by a garage mechanic in Nebraska (in a remark conveyed to me by R. L. Hall): "A teaspoon of sewage would spoil a barrel of wine, but a teaspoon of wine would do nothing at all for a barrel of sewage." What we might call the positive transvaluation of food does occur in a more salient way in other cultures. The clearest example comes from the work of Meigs (1978, 1983) on the Hua of New Guinea. For these people, the vital essence of a person, or nu, extends into all of the things he or she interacts with. Hence, the food one has grown, shot, killed, or cooked contains some of one's vital essence (in the sense that we might say one's fingernail parings are part of the self). For the Hua, food that contains the nu of those in a positive relation to the eater (e. g., for a male initiate, elder males) can nourish, while food that contains the nu of those in a potentially hostile relation to the eater (e. g., for the male initiate, mature females) can harm. Here we have, in a potent and salient way, something like positive contamination. The Hua conception that such positively or negatively contaminated foods are beneficial or harmful may well be masking or coexisting with a "positive disgust" response to them. The general suggestion is that with the possible exception of a few core disgusts (e. g., feces), disgust is a context-sensitive response (Meigs, 1978, 1983). Regardless of the status of positively transvalued food, the weakening of the negative bias in humans is clear. I would suggest that the gain in strength of the positive system in humans is related to culture (P. Rozin 1982). Enculturation amounts to the learning of many rules, at least some of which become internalized as positive and negative values. What better way to ensure enculturation than to have the culture members want (like) what the culture values, and dislike what it rejects? Such internalization obviates the necessity for rules and proscriptions to ensure compliance. It is possible that the unique presence of both culture and plentiful acquired likes in humans is not accidental.

,Sources of Variance in Food Preferences and Attitudes
The major determinant of what is eaten is probably availability, a factor that is to a large extent independent of any individual's behavior. Economic factors also severely constrain food choice. However, working from a psychological perspective, I will focus on differences among people in the absence of availability and economic constraints. The determinants of individual differences in food choice can be categorized as biological, cultural, or individual (psychological). If one was interested in determining as much as possible about an adult's food preferences and attitudes, and could only ask him one question, the question should undoubtedly


III. Nutritional and Biopsychological Constraints

be: what is your culture or ethnic group? But what is responsible for the substantial differences in food preferences and attitudes among members of the same culture? Surprisingly, we do not know. There is not a very informative second question to ask our hypothetical person. Sex and biological factors such as differences in taste sensitivity account for very little variance. The family should be a potent force. But studies on parent-child resemblance in food preferences (within a more or less homogeneous culture) have found either no relation or a very small one (correlations usually below 0.3) between children's preferences and those of their parents (Birch 1980; Pliner 1983; Rozin, Fallon, and Mandell 1984). The only substantial within-culture parent-child resemblance that we know of has to do with a particular food attitude: disgust sensitivity-the degree of concern for cleanliness of foods and the possibility that they have contacted offensive substances (Rozin et al. 1984). Although family resemblance can be caused by genetic factors or experience, the meager evidence we have implicates experience as the more important factor. For example, mother-father correlations in preferences or attitudes are equal to or higher than parent-child correlations (see, Rozin, Fallon and Mandell 1984), and heritabilities for taste preferences are very low (Greene, Desor, and Maller 1975). However, there are well-documented genetically based differences in sensitivity to some bitter compounds in humans (Fischer et al. 1961), and these show some weak relationship to preferences for bitter foods. Metabolic differences among individuals and ethnic groups may account for different preferences in some individuals, and some differences in cultural practices (Katz 1982; Simoons 1982). A case in point is lactose intolerance. In this case, a genetically determined metabolic difference sets up different contingencies related to eating (the presence of gastrointestinal distress after drinking milk), and this consequence influences food preferences through learning mechanisms that I have already discussed.

Biology and Culture as Interacting Constraints
We can view biological and cultural factors as establishing constraints or predispositions within which any individual develops a particular unique set of preferences and attitudes. We have discussed some of the factors accounting for these individual differences in terms of mechanisms for producing likes or dislikes for foods. The psychological (individual), cultural, and biological domains interact in complex ways. In particular, biological factors are expressed in individuals, and culture is created by individuals (see P. Rozin 1982 for a detailed treatment). One might reasonably expect biological and behavioral features of


7. Psychological Perspectives on Food Preferences and Avoidances

human food selection to be represented in cultures, and they are. (I leave aside the ways in which culture represents and adapts to nutritional as opposed to behavioral universals or individual differences; see Harris 1974, Katz 1982; or Simoons 1982 for examples of this approach.) There are rather few examples of the representation of genetically based and food-related behavioral biases in cultural institutions, largely because there are very few biological biases. However, the extraordinary popularity of sweets worldwide is represented in many cultural institutions and can be traced to a simple, genetically based preference for sweet substances. The domestication of plants producing sweet edibles and ultimately sugar refining itself, the colonial slave system, and the development of artificial sweeteners are all instances of cultural institutions prompted by our basic urge for sweets (Mintz 1985; P. Rozin 1982). Similar arguments can be made for avoidance (at the level of culture) of bitter foods, or removal of the bitterness. However, there are many counterinstances, such as the widespread ingestion by humans of foods that taste bad innately (coffee, tobacco, alcohol, chili pepper, etc.). Here, sociocultural forces arrange for a pattern of exposure to these initially undesirable items that ultimately induces a preference for them and reverses a biological bias (P. Rozin 1982).

The Psychological Importance of Flavorings: The Omnivore's Dilemma and Culinary Themes and Variations
Although flavorings make up a trivial portion of the human dietary in terms of calories or weight, their psychological importance is disproportionate. They account to a substantial degree for the distinctiveness of cuisines. Elisabeth Rozin has pointed out that most of the world's cuisines add characteristic flavor combinations called flavor principles to all of their staple foods (E. Rozin 1982, 1983). ("American" cuisine and some other temperate and Arctic cuisines do not have distinctive flavor principles.) Examples would be soy sauce, ginger root, and rice wine throughout China, chili pepper and tomato or lime in Mexico, and the spice combinations called "curries" in the different regions of India. The presence of these flavorings is essential to the satisfaction one gets from one's native cuisine; immigrant groups go to great lengths to obtain these flavorings (P. Rozin 1978). Flavor principles impart a characteristic flavor to most of the dishes in a cuisine. This culinary institution can be somewhat speculatively traced to a bit of human biology, the omnivore's dilemma. Fear and curiosity about new foods are represented in individual humans by a desire for familiarity in foods, an


III. Nutritional and Biopsychological Constraints

indicator of safety, and an opposing desire for variety and reaction against monotony in diets. Flavor principles may be a response to the desire for familiarity, since they give almost all foods a characteristic and familiar flavor. But what about the other side of the omnivore's dilemma, the desire for variety? Careful analysis indicates that the flavor principle of a cuisine is usually a group of flavorings that vary from dish to dish while maintaining a common identity or family resemblance. Thus, while chili pepper appears in almost all Mexican food, a wide variety of peppers with different flavors are used, and there is some systematic variation of these peppers from dish to dish. Similarly, the curries of any region of India are families of flavors that are mixed in varying ways in the menu cycle. In short, the practice of flavor principles can be seen to represent both sides of the omnivore's dilemma: variety within a general familiar constraint, or culinary themes and variations (E. Rozin and P. Rozin 1981; P. Rozin 1978). This may be a case where a culinary institution can be accounted for as a transformation of a basic bit of food selection biology.

Adaptation, Mechanism, and the Explanation of Culinary Practices
We move next to a home-base of "nutritional anthropology": the explanation of features of cuisine by explicating their adaptive value. As I have indicated, this enterprise has had its greatest successes in linking universal features of human nutrition (the need for protein, etc.), or genetically based metabolic differences among humans, with appropriately adapted culinary practices (see. e. g., Harris 1974; Katz 1982). A full explanation of these practices would require a historical reconstruction of how they came to be and an account of how they are passed along from generation to generation. This brings us to the contrast between adaptation and mechanism (see P. Rozin 1982 for a more extended discussion of this issue). The problem is: to what extent is an adaptive explanation of a culinary practice a representation of the selective force that caused it to become part of a cuisine, and to what extent is it a representation of how the practice is justified by contemporary users? Let us consider the latter question first. For some adaptive culinary practices, it is likely that the current motivation (psychology) overlaps the adaptive value. For example, the leaching of cyanide from bitter manioc may well be understood for what it is by contemporary users: the effects of the cyanide are rapid and striking. On the other hand, the elaborate and adaptive tortilla technology (Katz, Hediger, and Valleroy 1974) may be another story. The nutritional benefits of this technology, including the production of a better amino acid balance in the com, are not obviously perceptible in the nutritional effects of the food over short periods of time. In


7. Psychological Perspectives on Food Preferences and Avoidances

fact, I asked a number of women in a Mexican highland village why they boil their com with lime (calcium hydroxide) (P. Rozin 1982). The uniform response was that it was hard to roll out a tortilla made with com cooked without lime. This, of course, is a much more concrete and observable consequence of the technology. The question is, what is the relationship between the adaptive explanation and the explanation of the contemporary user, and which of these (if not both) is involved in the adoption of the technique in the culture? The same question arises for the nutritionally and medically beneficial effects of cooking, as opposed to the fact that it makes food easier to chew and better tasting, and for the relationship between the vitamin content of chili pepper and the fact that it facilitates the chewing of mealy diets. The point is simply that we have an interesting problem in attempting to relate some individual, psychological, motivational, and mechanistic explanations with adaptive ones.

Psychology, Anthropology, and Reconstructing History
In an attempt to explain culinary practices, the hardest step is to account for origins and early stages. (For example, how were the first steps in the tortilla technology discovered and institutionalized?) The problem is, of course, that we rarely have records of such events. Seemingly insignificant chance events, as in evolution, can have far-reaching consequences. For instance, I have speculated that the history of corn in Western Europe might have been very different if Cortez had had a European woman in his party (P. Rozin 1982). What was brought back to the Old World was a collection of com kernels, with perhaps some very, very stale tortillas. But Mexicans eat their com in the form of the highly palatable (fresh) tortilla, made with a technology practiced almost exclusively by women. It is likely that none of Cortez's men took the time to learn the technology, as a woman would have, and that the technology was never established in Europe. Therefore, a highly palatable form of com did not have a chance to take root on the continent. The frustrations of dealing with such problems are not unique to the nutritional anthropologist. The nutritional psychologist interested in accounting for an individual's preferences faces the same problem. Experiences like taste aversions (eating a food and getting sick) are often unique, one-time occurrences and often cannot be traced. In fact, we cannot account for most withinculture variation and must consider the possibility that many preferences are caused by unique and/or chance events. Our only solace is that, in contrast to the situation of the culture historian or anthropologist, the history of an individual's food preferences is a part of his past and is locked somehwere in his head-and perhaps as well in the heads of some members of his family. It is


III. Nutritional and Biopsychological Constraints

likely that anthropologists and psychologists will both have a great deal of trouble in excavating these histories. Fortunately, both disciplines have many other very important things to do, and hopefully they will do some of them together.

This paper is, in large part, constituted from three earlier publications by the author (Rozin 1982, 1983; Rozin and Fallon 1981). The author thanks April Fallon, Marcia Pelchat, Patricia Pliner, and Elisabeth Rozin for their contributions to the research discussed and for comments on the three papers mentioned above. The research described here was supported by National Science Foundation Grant BNS 76 80108 and National Institutes of Health Grant HD 12674.

References Cited
Angyal, A. 1941 Disgust and Related Aversions. Journal ofAbnormal and Social Psychol-

ogy 36: 393-412.
Appadurai, A. 1981 Gastropolitics in Hindu South Asia. American Ethnologist 8:494-511. Barker, L. M. (ed.) 1982 Psychobiology ofHuman Food Selection. Westport, Conn.: AVI. Beauchamp, G. K. 1981 Ontogenesis of Taste Preferences. In Food, Nutrition, and Evolution, D. Walcher and N. Kretchmer, eds. New York: Masson. Beauchamp, G. K.; M. Bertino; and K. Engelman 1983 Modification of Salt Taste. Annals ofInternal Medicine 98:763-69. Bern, D. 1967 Self-Perception: An Alternative Interpretation of Cognitive Dissonance Phenomena. Psychological Review 74: 183-200. Birch, L. L. 1980 The Relationship Between Children's Food Preferences and Those of Their Parents. Journal ofNutrition Education 12: 14-18. Birch, L. L.; D. Birch; D. W. Marlin; and L. Kramer 1982 Effects of Instrumental Consumption on Children's Food Preference. Appetite 3: 125-34. Birch, L. L.; S. I. Zimmerman; and H. Hind 1980 The Influence of Social-Affective Context on the Fonnation of Children's Food Preferences. Child Development 51: 856-61.


7. Psychological Perspectives on Food Preferences and Avoidances

Nonnal Control of Omnivore Intake by Taste and Smell. In The Determination of Behavior by Chemical Stimuli, ]. Steiner, ed., pp. 233-43. London: Infonnation Retrieval. Booth, D. A.; P. Mather; and]. Fuller 1982 Starch Content of Ordinary Foods Associatively Conditions Human Appetite and Satiation. Indexed by Intake and Eating Pleasantness of Starch-Paired Flavors. Appetite 3: 163-84. Cines, B., and P. Rozin 1982 Some Aspects of the Liking for Hot Coffee and Coffee Flavor. Appetite 3:23-34. Cowart, B.]. 1981 Development of Taste Perception in Humans: Sensitivity and Preference Throughout the Life Span. Psychological Bulletin 90: 43-73. Davis, C. Self-selection of Diets by Newly-weaned Infants. American Journal of 1928 Diseases of Children 36: 651-79. Douglas, M.

Booth, D. A. 1982


Purity and Danger: An Analysis of Concepts of Pollution and Taboo.

London: Routledge and Kegan Paul. Fallon, A. E., and P. Rozin The Psychological Bases of Food Rejections by Humans. Ecology ofFood 1983 and Nutrition 13: 15-26. Fallon, A. E.; P. Rozin; and P. Pliner The Child's Conception of Food: The Development of Food Rejections, 1984 with Special Reference to Disgust and Contamination Sensitivity. Child Development 55: 566-75. Ferro-Luzzi, G. E. 1975 Food Avoidances of Indian Tribes. Anthropos 70: 387-427. Fischer, R.; F. Griffin; S. England; and S. M. Gam 1961 Taste Thresholds and Food Dislikes. Nature 191: 1328. Frazer, ]. G. 1959 The New Golden Bough: A Study in Magic and Religion (abridged). New York: Macmillan. Garb, ]. L., and A. Stunkard 1974 Taste Aversions in Man. American Journal ofPsychiatry 131:1204-7. Goldstein, A., and S. Kaizer 1969 Psychotropic Effects of Caffeine in Man, III: A Questionnaire Survey of Coffee Drinking and Its Effects in a Group of Housewives. Clinical

Pharmacology and Therapeutics 10:477-88.
Greene, L.; ]. A. Desor; and O. Maller 1975 Heredity and Experience: Their Relative Importance in the Development of Taste Preferences. Journal of Comparative and Physiological Psychology 89: 279-84.


III. Nutritional and Biopsychological Constraints

Harris, M. 1974 Katz, S. 1982

Cows, Pigs, Wars and Witches: The Riddles ofCulture. New York: Random House. Food, Behavior and Biocultural Evolution. In The Psychobiology of Human Food Selection, L. M. Barker, ed., pp. 171-88. Westport,

Conn.: AV1. Katz, S.; M. L. Hediger; and L. A. Valleroy 1974 Traditional Maize Processing Techniques in the New World: Anthropological and Nutritional Significance. Science 184: 765-73. Krondl, M., and D. Lau 1982 Social Detenninants in Human Food Selection. In Psychobiology of Human Food Selection, L. M. Barker, ed., pp. 139-51. Westport, Conn.: AVI. Lepper, M. R. 1980 Intrinsic and Extrinsic Motivation in Children: Detrimental Effects of Superfluous Social Controls. In Minnesota Symposium on Child Psychology, vol. 14, W. A. Collins, ed., pp. 155-214. Hillsdale, N.].: Lawrence Erlbaum. Logue, A. W.; 1. Ophir; and K. F. Strauss 1981 The Acquisition of Taste Aversions in Humans. Behavior Research and Therapy 19: 319-33. Maga,]. A. Capsicum. Critical Reviews ofFood Science and Nutrition 7: 177-99. 1975 Meigs, A. S. A Papuan Perspective on Pollution. Man 13:304-18. 1978 Meigs, A. S. Food, Sex, and Pollution: A New Guinea Religion. New Brunswick, 1984 N.].: Rutgers University Press. Mintz, S. Sweetness and Power: The Place ofSugar in Modern History. New York: 1985 Viking. Ortner, S. B. 1973 Sherpa Purity. American Anthropologist 75: 49-63. Pelchat, M. L., and P. Rozin 1982 The Special Role of Nausea in the Acquisition of Food Dislikes by Humans. Appetite 3:341-51. Pliner, P. The Effects of Mere Exposure on Liking for Edible Substances. Appetite 1982 3:283-90. Pliner, P. 1983 Family Resemblance in Food Preferences. Journal of Nutrition Education 15: 137-40. Pliner, P.; P. Rozin; M. Cooper; and G. Woody 1985 The Minimal Role of Specific Post-Ingestional Effects in the Acquisition of Liking for Foods. Appetite 6:243-52.


7. Psychological Perspectives on Food Preferences and Avoidances

Richter, C. P.

1943 1956

Total Self-Regulating Functions in Animals and Human Beings. Harvey Lecture Series 38:63-103. Salt Appetite of Mammals: Its Dependence on Instinct and Metabolism. In L'Instinct Dans Le Comportment des Animaux et de L'Homme, Fondation Singer Polignac, ed., pp. 577-629. Paris: Masson.
The Structure of Cuisine. In The Psychobiology of Human Food Selection, L. M. Barker, ed., pp. 189-203. Westport, Conn.: AVI.

Rozin, E.

Rozin, E.


Ethnic Cuisine: The Flavor Principle Cookbook. Brattleboro, Vt.: Ste-

phen Greene. Rozin, E., and P. Rozin 1981 Culinary Themes and Variations. Natural History 90(2):6-14. Rozin, P. The Selection of Food by Rats, Humans and Other Animals. InAdvances 1976 in the Study ofBehavior, vol. 6, ]. Rosenblatt, R. A. Hinde, C. Beer, and E. Shaw, eds., pp. 21-76. New York: Academic Press. Rozin, P. 1978 The Use of Characteristic Flavorings in Human Culinary Practice. In Flavor: Its Chemical, Behavioral and Commercial Aspects, C. M. Apt, ed., pp. 101-27. Boulder, Colo.: Westview. Rozin, P. 1979 Preference and Affect in Food Selection. In Preference Behavior and ChemorecejJtion, ]. H. A. Kroeze, ed., pp. 289-302. London: Information Retrieval. Rozin, P. Human Food Selection: The Interaction of Biology, Culture and Indi1982 vidual Experience. In The Psychobiology ofHuman Food Selection, L. M. Barker, ed., pp. 225-54. Westport, Conn.: AVI. Rozin, P. 1984 The Acquisition of Food Habits and Preferences. In Behavioral Health: A Handbook of Health Enhancement and Disease Prevention. ]. D. Matarazzo et aI., eds., pp. 590-607. New York: John Wiley. Rozin, P., and B. M. Cines 1981 Multiple Motives for Drinking Coffee. Unpublished Manuscript. 1982 Ethnic Differences in Coffee Use and Attitudes to Coffee. Ecology of

Food and Nutrition 12: 79-88.
Rozin, P.; I. Ebert; and]. Schull 1982 Some Like It Hot: A Temporal Analysis of Hedonic Responses to Chili Pepper. Appetite 3: 13-22. Rozin, P., and A. E. Fallon 1980 The Psychological Categorization of Foods and Non-Foods: A Preliminary Taxonomy of Food Rejections. Appetite 1: 193-20l. Rozin, P., and A. E. Fallon 1981 The Acquisition of Likes and Dislikes for Foods. In Criteria of Food


III. Nutritional and Biopsychological Constraints

1986 Rozin, P.; A. E. Fallon; and R. Mandell 1984 Family Resemblances in Attitudes to Foods. Developmental Psychology 20:309-14. Rozin, P.; I. Gruss; and G. Berk 1979 The Reversal of Innate Aversions: Attempts to Induce a Preference for Chili Pepper in Rats. Journal of Comparative and Physiological Psychology 93: 1001-14. Rozin, P., and]. W. Kalat 1971 Specific Hungers and Poison Avoidance as Adaptive Specializations of Learning. Psychological Review 78:459-86. Rozin, P., and K. Kennel 1983 Acquired Preferences for Piquant Foods by Chimpanzees. Appetite, 4:69-77. Rozin, P.; L. Millman; and C. Nemeroff 1986 Operation of the Laws of Sympathetic Magic in Disgust and Other Domains. Journal ofPersonality and Social Psychology 50: 703-12. Rozin, P., and D. Schiller 1980 The Nature and Acquisition of a Preference for Chili Pepper by Humans. Motivation and Emotion 4:77-10l. Senn, M. ]. E., and A. ]. Solnit Problems in Child Behavior and Development. Philadelphia: Lea & 1968 Febiger. Simoons, F. ]. 1982 Geography and Genetics as Factors in the Psychobiology of Human Food Selection. In The Psychobiology of Human Food Selection, L. M. Barker, ed., pp. 205-24. Westport, Conn.: AVI. Solomon, R. L. 1980 The Opponent-Process Theory of Acquired Motivation. American Psychologist 35: 691-712. Stein, M.; P. Ottenberg; and N. Roulet 1958 A Study of the Development of Olfactory Preferences. American Medical Association Archives ofNeurology and Psychiatry 80: 264-66. Steiner, ]. Human Facial Expressions in Response to Taste and Smell Stimulation. 1979 Advances in Child Development and Behavior 13:257-95. Tambiah, S. ]. Animals Are Good to Think and Good to Prohibit. Ethnology 8:423-59. 1969 Zahorik, D. Learned Changes in Preferences for Chemical Stimuli: Asymmetrical 1979 Effects of Positive and Negative Consequences, and Species Differences in Learning. InPre/erence Behaviour and Chemoreception,]. H. A. Kroeze, ed., pp. 233-46. London: Information Retrieval.

Acceptance: How Man Chooses What He Eats, ]. Solms and R. L. Hall, eds., pp. 35-48. Zurich: Forster. A Perspective On Disgust. Psychological Review (in press).


7. Psychological Perspectives on Food Preferences and Avoidances

Zajonc, R. B. 1968

Attitudinal Effects of Mere Exposure. Journal ofPersonality and Social

Psychology 9 (Part 2): 1-27.
Zellner, D. A.; P. Rozin; M. Aron; and C. Kulish 1983 Enhancement of Human's Liking for a Flavor by Pavlovian Processes. Learning and Motivation 14: 388-50.



The Preference for Animal Protein and Fat: A Cross-Cultural Survey
FROM BOTH A DIACHRONIC AND A SYNCRHONIC PERSPECTIVE, the preference for some type of animal protein and animal fat in human diets is a cultural imperative (Abrams 1979). Usually animal fat accompanies animal protein in foods (USDA 1975). Hence, when humans consume animal proteins, they also automatically consume some animal fats (although at certain seasons of the year non-domesticated animal species may be a poor source of fat). The origin of the human preference for some type of animal protein and animal fat may be found in the evolutionary antecedents of the species. Although predominantly vegetarian, most contemporary primate species include in their diet some animal protein from such sources as small animals, birds, eggs, insects, and lizards. Even primates that do not purposely seek animal protein may eat small amounts of it adventitiously in the form of insects, insect eggs, and larvae that are present in consumed fruits and plants (S. Campbell 1978; Eimerl, DeVore, et a1. 1965; Galdikas-Brindamour 1975; Harding 1981; Perry 1976; Starin 1981; Van Lawick-Goodall 1971). Observed baboons and chimpanzees have been found to hunt, often cooperatively, for small animals such as monkeys, baby gazelles, and other young ungulates, and to share the bounty. In contrast, joint search for or sharing of abundant plant food would be unusual (Eimerl, DeVore, et a1. 1965; Hamilton 1978; Harding and Strum 1976; Strum 1981; Teleki 1973, 1981; Van Lawick-Gooda1l1971). Uncertainty still exists regarding specific features of the social organization and food habits of the australopithecines and Homo habilis. However, there is general agreement that the earliest humans were omnivorous (Dart 1969; Johanson and White 1979; Potts and Shipman 1981; Zihhnan 1983). The ability of early hominids to survive through a hunting-gathering-scavenging mode of subsistence has been tested by experiments carried out in the australopithecines' ancient habitat, in conditions thought to resemble those of the early Pleistocene. Schaller and Lowther (1969), traveling without modem weapons, were


III. Nutritional and Biopsychological Constraints

able to secure ample meat supplies. During the course of several days and over many miles, they stumbled upon two gazelle carcasses. They encountered a cheetah making a kill, and they could easily have partaken of the prey. They found bits of brain remaining from another kill that had been almost entirely consumed. In addition, they were able to secure many small animals. In all, they found more than 655 pounds of meat. The case for meat eating by early humans or proto-humans is further strengthened by evidence that between the late Pliocene (three million years ago) and the mid-Pleistocene (one million years ago), there was a significant reduction in the varieties of carnivorous species inhabiting southern Africa. It is suggested that the decrease was due, at least partially, to the arrival and subsequent evolutionary success of meat-eating hominids (Klein 1977). A clue as to the diet of the australopithecines can be deduced from that of the contemporary Australian Aborigine, who eats whatever edible food is available in the natural habitat. In addition to plant food, the selections range from worms and insects to the six-foot-high kangaroo (Birdsell 1972; D. Harris, 1978; Chapter 14). It has been postulated that meat played an important role in the development of Homo sapiens (Washburn 1961). By consuming animals that live on plant foods, humans obtained a highly concentrated and complete form of protein that contains all of the essential amino acids, converted from plant into animal protein (White et al. 1973). Animal protein is far more complete and concentrated than plant protein. In addition, animal protein reduces the bulk needed for energy requirements and reduces by two-thirds the time necessary for eating. More time could thus be invested in hunting, especially in the pursuit of larger animals, which required higher levels of cooperation and the development through natural selection of an increased ability to postpone immediate desire (Bronowski 1972). The selective adaptations that resulted in the emergence of Homo erectus included the hunt for wild game. Homo erectus developed specialized tools and, as a skillful hunter, was able to spread from tropical Africa into Europe and Asia, the availability of game being relatively constant from season to season in temperate regions, unlike the situation with plant food (B. Campbell 1976; Chard 1975). The driving force that led Cro-Magnon to occupy all of the world's continents except Antarctica was the constant quest for game. The modern Lapps apparently follow reindeer herds much as early man followed game (Bronowski 1972). Diverse archaeological excavations reveal that Homo erectus had mastered the hunting of large game (Butzer 1971, 1977). Examples include finds at Ambrona and Torralba in Spain (White et al. 1973), Terra Amata in France (Lumley 1972), Olduvai Gorge in Africa (Leakey 1971), and Choukoutien in China (Braidwood 1975; Treistman 1972). With the advent of Homo sapiens, the


8. The Preference for Animal Protein and Fat

quest for game became more efficient, perhaps so efficient that many large species were predated to the point of extinction (Martin 1967). The central theme of the paleolithic food quest was the hunt for meat (Grivetti 1978). Both archaeological finds (B. Campbell 1976; Solecki 1971) and cave art (Gotham 1950) reflect the preoccupation with the quest for game. The debate may continue about the extent to which hunters and gatherers depended on animal protein. Most likely the plant-animal ratio varied. However, evidence seems to indicate a distinct preference for animal protein in the paleolithic hunting and gathering economies (Harding and Teleki 1981; Lee and DeVore 1968). Current paleoanthropological data indicate that hominids have existed for at least three to four million years Oohanson and White 1979). For more than 99 percent of this time span, hominid societies depended on hunting and gathering food. This period came to an end only in the last ten thousand years, with the adoption of agricultural modes of production. Meat had been the mainstay for paleolithic humankind. Due to growing population pressure, grains became the staff of life in horticultural and agricultural societies (Cohen 1977). Nevertheless, in virtually all societies, meat and other animal products remain preferred foods. In most of the contemporary world, when people entertain guests for a meal, meat or some form of animal protein is usually featured as an expression of cordiality and friendship. According to Lee and DeVore (1968), the hunting and gathering lifestyle represents the most successful and enduring adaptation ever achieved by humankind. Hunters and gatherers probably possessed the knowledge that plants grow from seeds from earliest times (Cohen 1977). No known human population is ignorant of the relationship between plants and the seeds from which they grow (Flannery 1968). Bronson (1975) maintains that activities resembling cultivation existed as early as the Pleistocene period. The growing of useful plants was neither a unique nor a revolutionary event of recent origin. Murdock's (1934) and Service's (1958, 1971, 1978) profiles of non-industrial societies all reflect a preference for animal protein, albeit protein of various types. Among the Tasmanians, for example, although fish was taboo, the consumption of oyster, crab, and other shellfish was favored. In addition to ants and grubworms, snake, lizard, kangaroo, wallaby, bandicoot, and wombat were relished (Basedow 1925; Bourne 1953; Murdock 1934). Australian Aborigines ate practically any animal they could find. They especially prized kangaroo and other marsupials, and also ate grubworms, which have a high fat content. The Yahgan of Tierra del Fuego favor seal and whale and subsist largely on mussel, conch, shellfish, and limpet (Basedow 1925; Bourne 1953; Cooper 1946; Mountford 1974; Tonkinson 1978; Warner 1958). Andaman Islanders prefer pig, civet cat, large lizard, snake, rat, dugong, turtle, fish crab, crayfish, and


III. Nutritional and Biopsychological Constraints

mollusk. The distribution and eating of turtle and pig are important in their ceremonies. The turtle is highly valued for its fat. A young bachelor who kills a pig must distribute the choicest parts to the elders. This illustrates their preference for pork, which is also high in fat (Radcliffe-Brown 1964; Service 1971). The Dobe-area Bushmen of the Kalahari desert consider 54 species of animals edible, and relish 17 of them as choice (Lee 1968). In other areas, Bushmen include many more animals in their diet (Bleek 1928; Knobel 1952; Lee 1972, 1974; Marshall 1965; Silberbauer 1981). The effort expended to obtain a certain amount of animal food reflects a strong preference for it. Traditionally the Eskimos subsisted almost entirely on the flesh of animals such as caribou, seal, fish, and bird (Stefansson 1957, 1960). The mainstay of the Lapp diet was the flesh and milk of reindeer (Bronowski 1972; Weyer 1961). Although the Jivaro of the Amazon jungle depend primarily on crops, game is higWy prized. The Jivaro actively hunt for wild hog, peccary, bird, crocodile, tortoise, snail, snake, lizard, and frog, as well as ant larvae for their fat content (Service 1958; Steward and Metraux 1948). Among the Mehinaku of the Brazilian Amazon jungle, fish is the mainstay of the diet, but meat from wild animals is higWy cherished (Gregor 1981). Fishing is of prime importance to the Kalapalo of central Brazil (Basso 1973). Because of the Kalapalo taboo against killing most animals, few are eaten. The Yanomamo include monkey, wild turkey, wild pig, armadillo, anteater, tapir, deer, alligator, rodents, and several bird species among their valued foods (Chagnon 1968). Andean and Meso-American staples were plant foods such as maize, white potato, beans, and squash. Animals that might have been suitable for domestication had been hunted to extinction as these societies were beginning to develop. The Andean civilizations domesticated the llama and alpaca. But these creatures, valuable as pack aniamls as well as for their wool, were rarely used as food. The Andeans domesticated the muscovy duck, dog, and guinea pig and ate fish and wild animals. All were highly regarded but, being scarce, were limited to festive occasions (Rowe 1946; Service 1978). In Meso-American cultures, exemplified by the Aztecs, only two animals were domesticated: the Mexican hairless dog and the turkey. Both were highly valued (Duran 1964; Peterson 1959; Sahagun 1963). Harner (1977) maintains that in order to augment their meager supply of animal foods, the Aztecs turned to cannibalism, legitimated through religious sanction as a sacrifice to the gods. The Aztecs ate ants, mosquito eggs, toasted grasshoppers, and many types of worm, including the well-known maguey worm. They used the stingless honey bee for its honey. Such items are still prized and eaten in many Mexican Indian villages (D. Heyden, pers. comm.; Taylor 1975). Predominantly horticultural societies esteem animal proteins and animal fats, although the types consumed may vary in different cultures because of ecologi-


8. The Preference for Animal Protein and Fat

cal factors and cultural values. For example, the Ganda of Uganda, a horticultural people of east Central Africa, supplement their predominantly plant food diet with wild game, such as antelope, buffalo, elephant, and smaller animals. They also keep chickens, goats, sheep, and cattle. Their cows yield little milk and its consumption is surrounded by many taboos. Barren cows and bulls may be slaughtered for food. Animals that die from natural causes may also be eaten. Animal protein would be eaten more frequently if it were available (Murdock 1934; Roscoe 1911). In contrast to the Ganda, the horticultural Dahomeans of West Africa engage in fishing and hunting of wild game. These animal protein sources are supplemented with pig and poultry (Murdock 1934, 1959). The subcontinent of India is usually characterized as predominantly vegetarian. Despite this view, and the diversity of the diets of various religious and tribal groups, some types of animal proteins and animal fats are preferred foods (see Chapter 18). The Hindus, who constitute the overwhelming majority of the population, universally reject beef. Yet all Hindus regard milk and ghee (clarified butter) as preferred foods. Moreover, the Hindus eat certain types of meat. The lower their caste, the less frequently they can afford to include costly meat in their diet (HRAF 1964; Moore 1970; Vaidyanathan 1983; and see Chapter 18). Agricultural societies differ from horticultural ones in that they utilize draft animals in farming. Frequently animals that are no longer useful in the field are eaten. The Kalinga rice-farmers of the Philippines eat the water buffalo after it is too old for plowing. The Kalinga value the domesticated pig, beef animal, and fowl (Service 1958). In contrast, the Aymara of Bolivia supplement their basic potato diet with fish, pork, mutton, guinea pig, and eggs only during festivities (Buechler 1971). In Morocco, which is typical of Middle Eastern farming societies, mutton and goat are highly esteemed foods, and beef, chicken, and milk are favored (Service 1958). Generally in pastoral and nomadic herding societies the diet consists mainly of animal foods. The Kazaks of Central Asia subsist almost entirely on their sheep and goat herds. Their prime summertime sustenance is provided by milk from the goat, cow, ewe, mare, and camel (Murdock 1934). The culture of the Todas in southern India revolves around the water buffalo. The Todas subsist almost entirely on dairy products derived from buffalo milk, along with some plant foods (Murdock 1934; Rivers 1906). Among the Nuer of East Africa, milk, blood, and meat from their cattle provide the main source of food. Milk is their year-round staple (Evans-Pritchard 1940). For the nomadic Rwala Bedouins of Arabia, camel milk is the staple food (Service 1978). A survey of the autochthonous Amerindian societies of North America reveals that all cultures preferred some type of animal protein and animal fat. The Eskimos, the Northwest Coast Amerindians, and the Plains societies all subsisted primarily on animal protein and animal fat. In the Southwest, where


III. Nutritional and Biopsychological Constraints

these were scarce, cultures such as the Hopi and Zuni consumed little animal protein and animal fat. The Navaho obtained more by hunting wild game. The Amerindian societies of the Eastern Woodlands cultural areas valued wild game such as deer, bear, and smaller mammals, as well as fowl, fish, shellfish, and fresh water mollusk (Owen, Deetz, and Fisher 1967; Spencer, Jennings, et aI. 1965). The basic diet of the contemporary Navaho is bread and mutton or goat flesh (Kluckhohn and Leighton 1974). Often the near depletion of game by one Amerindian tribe in an area has led to incursions into the hunting areas of others. The wars between the Chippewa and Dakota typify the conflicts that ensued (Hickerson 1962). A survey of the indigenous cultures of South American reveals a similar pattern, with all cultures using and prizing some animal protein in the diet. The range runs the gamut from an almost totally animal diet, as in the Tierra del Fuego area, to the use of small amounts of animal protein, as among the people of the Andean cultural area (Lyon 1974; Steward and Faron 1959; Steward and Metraux 1948). Autochthonous cultures in Africa reveal a similar pattern. The type and amount of protein and fat consumed correlates with ecological and subsistence patterns ranging from those of the Pygmies and Bushmen to those of horticultural and pastoral or herding societies (Moore 1970; Murdock 1959; Ottenberg 1960). The cultures of Asia and Oceania reflect similar patterns (Moore 1970). Apparently all contemporary pre-state peoples parallel the practice of prehistoric societies in devoting various amounts of their economic activity to the procurement of some type of animal protein and animal fat (Gabel 1967; Herskovits 1965). The importance of insects, spiders, and other small creatures in the diets of many cultures throughout the world is frequently overlooked. Such food is esteemed by many peoples and sometimes relished as a delicacy. These small animals augment the intake of other animal protein and animal fat. Insect foods consumed may include the egg, larva, pupa, and adult (Brothwell and Brothwell 1969; Myers 1982; Oliver 1967). Some of the small creatures eaten in various areas and by different societies have been noted by Bates (1959; 1967), Bodenheimer (1951), Bourne (1953), Bristowe (1932), Brothwell and Brothwell (1969), Curran (1937), Downs (1966), Essig (1934), Gibbs (1965), Gelfand (1971), Hallet (1966), D. Harris (1975, 1983), Hitchcock (1962), Hoffman (1947), Meyer-Rochow (1973), Pyne (1983), Quin (1959), Remington (1946), Ruddle (1973), Steward (1963a, 1963b), Taylor (1975), and Williston (1883a, 1883b). This research may be summarized as follows:
Ants: Australian Aborigines, Amerindian tribes (Digger, Onondaga, Jivaro, Miskito, Marquiritares, Yukpa), segments of populations (in Colombia, Thailand, Papua, West Irian, Tasmania, Zimbwawe, South Africa) Bees and wasps: China, Bunna, Malaya, Australian Aborigines, Sri Lanka,


8. The Preference for Animal Protein and Fat

Thailand, parts of Japan, Pedi of South Africa, Aun district in East Sudan, Khoisen people in South Mrica, Amerindian tribes (Waro, Chaco, Yanomami, Yukpa) Beetles: Papua, West Irian, Pedi of South Africa, Shona, Java, Sumatra, Burma, Malaya, Madagascar, New Zealand, 1urkey, Australian Aborigines Butterflies and moths: Eskimos, Amerindian tribes (Washo, Klamath, Modoc, Mono, Paiute), Australian Aborigines, Indonesia, Burma, Japan, China, Zambia, Congolese, Papua, West Irian, Polynesia, Madagascar, Shona, or Zimbwawe Centipedes: Guahibo Amerindians of Venezuela Cicadas: Australian Aborigines, Papua, West Irian, southern Thailand, ancient Greece Cockroaches: China, Thailand, Australian Aborigines, Bushmen of the Kalahari desert Crickets: Burma, Shone of Rhodesia (in times of famine), Papua, West Irian, Uganda, Thailand, Jamaican Amerindians, Japan (prior to use of insecticides) Dragonflies: Indonesia (Bali and Island of Lombok), Thailand (Ubon district), Nigeria, China, Papua, West Irian, Madagascar, Japan Flies: Paiute, Mono and Koso of the Sochone tribes, Washo, Modoc, Dog Rib of Canada, Nambicuara, Yukpa Praying Mantis and Walking Sticks: Papua, West Irian, China, Japan (prior to use of insecticides) Spiders: Papua, West Irian, Yanomami Amerindians, Diyarua Amerindians, Thailand, Burma, Annam, Cambodia, Maniana (south of Gambia), Madagascar, Bushmen Termites: Pemon Amerindians, Yanomami Amerindians, many Hindu groups in India, Indonesia, Thailand, Palawi, Rhodesia, Pgymies of Africa, Congo, Australian Aborigines, Burma, many Bantu tribes of South Africa

Extensive data on food types, prestige values, and preferences were collated in the Food Habits Survey of the Human Relations Area Files (HRAF 1964), which was commissioned by the U.S. Quartennaster Corps. The survey encompassed 383 cultures, of South Asia and East Asia (except China), Oceania, the Middle East, Africa, and Latin America. About a half-million informational items were gathered. Data were compiled for 38 types, categories, or aspects of foods, with an average of 35 characteristics for each culture. All societies consumed animal foods. As part of the study, these foods were ranked by order of occurrence, and their "acceptability" was rated according to the following scale: 1. Highly preferred: This category includes delicacies with great cultural value: for example, watermelons among the Aymara. Usually this rating does not apply to primary foods. 2. Pre/erred: These dishes are more higWy valued than most, but not necessarily favorites. 3. Accepted readily: Positive evaluations are noted in some written source. 4. Acceptable substitute: These dishes are not preferred but will be consumed in lieu of preferred dishes when the latter are not available.


III. Nutritional and Biopsychological Constraints

5. Disliked: Disfavor is expressed in some written source. 6. Variable: The preference for a dish may not be the same for different parts of the group, or the preference may vary from time to time. 7. No preferences: An author explicitly states that the people have no marked liking or dislike for a certain dish-for example, "They are indifferent" or "They can take it or leave it." 8. Other: A form of preference that cannot be classified in any of the above categories (HRAF 1964). Table 8.1 lists the most commonly consumed animals and animal products. Some hunting and gathering societies placed a premium on animal fats and marine oils. For example, the Eskimos utilized blubber (Stefansson 1960), and the Northwest Coast Amerindians used fish oil (Carbarino 1976; Service 1971). All horticultural and agricultural societies also favor oil or fat for use in food preparation-for example, in cooking, baking, or as a flavor enhancer, in spreads, and as cream in foods and in beverages. Tannahill's (1973) and Brothwell and Brothwell's (1969) histories of food demonstrate the important role of animal protein and animal fat throughout human history. Many contemporary cultures continue to use and to prefer animal fats, as demonstrated by BarerStein (1980) in her study of the following contemporary cultures: Argentina: Beef fat, butter Germany: Lard, bacon fat, rendered fat from geese, ducks, Australia: Butter (being replaced by margarine) chickens, horses Austria: Sour cream, whipped Greece: Butter cream, lard, butter Holland: Butter Brazil: Beef fat, butter, lard Hungary: Lard, butter Belgium: Butter, lard Iceland: Cream, butter, sheep Bulgaria: Butter fat, horse fat India: Butter, ghee China: Lard, strained pork suet I ran: Butter, fat from "fat-tailed for pastry sheep" Czechoslovakia: Butter, rendered pork and goose fat I reland: Bacon fat, butter Denmark: Butter Israel (and]ewish cooking generally): Butter Egypt: Butter England: Lard, butter Italy: Cream, butter Estonia, Latvia, and Lithuania: Korea: Fat eaten from pork cracklings and chicken skin Lard, bacon fat, butter, sour Lebanon: Clarified butter from cream, cream, whipped cream sheep or goat milk Finland: Butter, salt pork, bacon fat Malta: Lard, butter France: Butter, cream, lard, Mexico: Lard, some butter chicken and goose fat Morocco: Clarified butter


8. The Preference for Animal Protein and Fat

TABLE 8.1. Commonly Consumed Animals and Animal Products

Animal Protein Type
Chicken (flesh and eggs) Cattle (meat and milk) Pig (domesticated) Goat Fish (not including specified small species) Sheep Duck (primarily flesh) Zebu, (primarily milk) Turtle (meat and eggs) Zebu (meat and milk) Dog Rat




166 159

2 2 2

67 49 46

3 3

42 42


aNumber of societies. bThe acceptability scale ranges from 1 ("highly preferred") to 5 ("disliked"). A score of 2 indicates that a food is preferred; a score of 3 indicates that it is accepted readily (HRA.F 1964). A rating of 1 ("highly preferred") is rarely applied to primary foods.

New Zealand: Butter, crean1,
lard, salt pork Norway: Butter, cream, sour cream, rendered duck and goose fat, lard, salt pork Philippines: Lard, butter Poland: Butter, lard, salt pork, bacon fat, rendered chicken, goose, and duck fat Portugal: Lard Romania: Lard, butter Russia: Butter Scotland: Butter, lard, suet Spain: Butter (mainly pork fat in Basque region) Sweden: Butter, lard, cream Switzerland: Butter, beef drip-

pIngs, chicken fat, lard, bacon fat Syria: Clarified butter from sheep or goat milk Turkey: Butter Ukraine: Lard, rendered salt pork, bacon ends, butter Uruguay: Beef fat, butter Vietna1'n: Pork fat Wales: Lard and Drippings, bacon fat, butter West Indies: Lard, butter, ghee

Yugoslavia (Croats, Serbs, Slovenes, Macedonians): Pork and
sheep fat, butter, cream, sour cream

Although environment is not a determining factor in diet, it is a limiting one. jill ecologically adaptive cultural relationship exists between the foods available


III. Nutritional and Biopsychological Constraints

and those selected. For various reasons, most societies place taboos on some types of foods. The prominence of animal food in the diet of a people is limited by its availability or scarcity. For example, the Hopi Indians, a group typical of the Southwestern cultures of the United States, were predominantly vegetarian because of the paucity of animal foods available to them. The rabbit, antelope, and prairie dog, which were available in limited numbers, contributed a small amount of eagerly sought animal protein and fat to the Hopi diet (Murdock 1934). At times, the general economic status of the population detennines how much animal protein and fat are eaten. In countries such as China, because of the subsistence-level income of the masses, little animal protein and fat may be consumed per capita. Although the affluent may indulge frequently in meat, fish, and eggs, the masses can afford such foods only occasionally (Wittfogel 1960). This pattern is found worldwide in varying degrees. In conclusion, cross-cultural data clearly demonstrate that all human cultures include some types of animal proteins and animal fats in their diets and esteem them higWy. Consumption of these foods is detennined, or limited, by ecological and economic factors. No exclusively vegetarian society has ever been discovered (Abrams 1980; Farb and Armelagos 1980). The preference for and use of animal protein and fat appears to have been a dominant human dietary trait from the earliest times. Both diachronic and synchronic cross-cultural data unequivocally reveal that animal protein and animal fat, as preferred foods, are universally present in human diets.

References Cited
Abrams, H. L. 1979 The Relevance of Paleolithic Diet in Determining Contemporary Nutritional Needs. Journal ofApplied Nutrition 31:34-59. 1980 Vegetarianism: An Anthropological/Nutritional Evaluation. Journal of Applied Nutrition 32: 53-87. Barer-Stein, T. 1980 You Eat What You Are, A Study of Ethnic Food Traditions. Toronto: McClelland and Stewart. Basedow, H. The Australian Aboriginal. Adelaide: F. W. Preece. 1925 Basso, E. B. The Kalapalo Indians of Central Brazil. New York: Holt, Rinehart and 1973 Winston. Bates, M. Insects in the Diet. American Scholar 29: 43-52. 1959 Gluttons and Libertines. New York: Random House. 1967


8. The Preference for Animal Protein and Fat

Birdsell, j. B. Human Evolution. Chicago: Rand McNally. 1972 Bleek, D. F. The Naron: A Bushman Tribe of the Central Kalahari. Cambridge: 1928 Cambridge University Press. Bodenheimer, F. S. 1951 Insects as Human Food. The Hague: Dr. W. Junk. Bourne, G. H. 1953 The Food of the Australian Aboriginal. Proceedings of the Nutrition Society 12: 58-65. Braidwood, R. j. 1975 Prehistoric Men. Glenville, Ill.: Scott Foresman. Bristowe, W. S. 1932 Insects and Other Invertebrates for Human Consumption in Siam. Transactions of the Entomology Society (London) 80: 387-404. Bronowski, j. The Ascent of Man. Boston: Little, Brown. 1972 Bronson, B. The Earliest Fanning: Demography as Cause and Consequence. In Pop1975 ulation, Ecology and Social Evolution, Steven Polgar, ed., pp. 53-78. The Hague: Mouton. Brothwell, D., and P. Brothwell 1969 Food in Antiquity. London: Thames and Hudson. Buechler, H. C., and j. M. Buechler 1971 The Bolivian Aymara. New York: Holt, Rinehart and Winston. Butzer, K. W. 1971 Environment and Archaeology: An Ecological Approach to Prehistory. Chicago: Aldine. 1977 Environment, Culture and Human Evolution. American Scientist 65: 576. Campbell, B. G. 1976 Humankind Emerging. Boston: Little, Brown. Campbell, S. 1978 Noah's Ark in Tomorrow's Zoo: Animals Are A-comin', Two by Two. Smithsonian 8: 42-50. Carbarino, M. S. Native American Heritage. Boston: Little, Brown. 1976 Chagnon, N. A. 1968 Yanomamo, The Fierce People. New York: Holt, Rinehart and Winston. Chard, P. S. 1975 Man in Prehistory. New York: McGraw-Hill. Cohen, M. N. 1977 The Food Crisis in Prehistory. New Haven: Yale University Press. Cooper, ]. M. 1946 The Yahgan. In Handbook of South American Indians. Bulletin 14, vol. 1, ]. H. Steward, ed. Bureau of American Ethnology. Washington, D. C.: Smithsonian Institution.


III. Nutritional and BiopsychologIcal Constraints

Curran, C. H. 1937 On Eating Insects. Natural History 43(2):84-89. Dart, R. 1969 Adventures with the Missing Link. New York: Viking. Downs, ]. F. 1966 The Two Worlds o/the Washo: An Indian Tribe o/California and Nevada. New York: Holt, Rinehart and Winston. Duran, Fray D. 1964 The Aztecs: The History of the Indies of New Spain. Translated with notes, D. Heyden and F. Horcasitas; Introduction, 1. Bernal. New York: Orion. Eimer!, S.; 1. DeVore; and the editors of Life 1965 The Primates. New York: Time. Essig, E. O. 1934 The Value of Insects to the California Indians. Science Monthly 38: 18186. Evans-Pritchard, E. E. 1940 The Nuer. Oxford: Oxford University Press. Farb, P., and G. Armelagos 1980 Consuming Passions: The Anthropology of Eating. Boston: Houghton Mifflin. Flannery, K. V. 1968 Archaeological Systems Theory and Early MesoAmerica. In Anthropological Archaeology in the Americas, B. Meggers, ed., pp. 67-87. Washington, D.C.: Anthropology Society of Washington. Gabel, C. 1967 Analysis ofPrehistoric Economic Patterns. New York: Holt, Rinehart and Winston. Galdikas-Brindamour, B. 1975 Search for the Great Apes. A National Geographic documentary film. Washington, D. C.: National Geographic. Gelfand, M. 1971 Diet and Tradition in an African Culture. London: E. and S. Livingstone. Gibbs, ]. L., Jr. 1965. Peoples ofAfrica. New York: Holt, Rinehart and Winston. Gotham 1950 Lascaux: Cradle 0/ Man's Art. A documentary film. Gotham. Gregor, T. 1981 We Are Mehinaku. A Public Broadcasting System documentary film. Boston: Public Broadcasting Associates. Grivetti, L. E. Culture, Diet and Nutrition: Selected Themes and Topics. Bioscience 1978 28(3): 171-77. Hallet, ]. 1966 Congo Kitabu. New York: Random House.


8. The Preference for Animal Protein and Fat

Hamilton, W. ].


Primate Carnivory and Its Significance to Human Diets. Bioscience

Harding, R. S. O. 1981 An Order of Omnivores: Nonhuman Primate Diets in the Wild. In Omnivorous Primates: Gathering and Hunting in Human Evolution. R. S. O. Harding and G. Teleki, eds., pp. 191-214. New York: Columbia University Press. Harding, R. S. 0., and S. C. Strum 1976 The Predatory Baboons of Kekopey. Natural History 85(3):46-53. Harding, R. S. 0., and G. Teleki (eds.) Omnivorous Primates: Gathering and Hunting in Human Evolution. 1981 New York: Columbia University Press. Harner, M. 1977 The Ecological Basis for Aztec Sacrifice. American Ethnologist 4: 117-

Harris, D. R. Adaptation to a Tropical Rain-Forest Environment: Aboriginal Subsistence in Northeastern Queensland." In Human Behaviour and Adaptation, N. Blurton-Jones and V. Reynolds, eds., pp. 113-34. London: Taylor and Francis. Herskovits, M. ]. The Economic Life ofPrimitive Peoples. New York: Norton. 1965 Hickerson, H. 1962 The Southwestern Chippewa: An Ethnohistorical Study. American Anthropological Association memoir no. 92, part 2, 64(3). Hitchcock, S. W. 1962 Insects and Indians ofthe Americas. Bulletin ofthe Entomological Society


ofAmerica 8: 181-87.
Hoffman, W. E. Insects as Human Food. Proceedings (of the Entomological Society of America) 49:233-37. Human Relations Area Files (HRAF) 1964 Food Habits Survey. New Haven: U.S. Quartennaster Corps. Johanson, D. C., and T. D. White 1979 A Systematic Assessment of Early African Hominids. Science 203:32130. Klein, R. G. 1977 The Ecology of Early Man in Southern Africa. Science 197: 115-26. Kluckhohn, C., and D. Leighton The Navaho. Cambridge: Harvard University Press. 1974 Knobel, L. 1952 Remnants of a Race. A Documentary film on the Bushmen of the Kalahari: Kalahari Films, Educational Distributors (P T D) Ltd., Encyclopedia Britannica Educational Corp.



III. Nutritional and Biopsychological Constraints

Leakey, M. D. 1971 Olduvai Gorge. Vol. 3. New York: Cambridge University Press. Lee, R. B. 1972 The !Kung Bushmen of Botswana. In Hunters and Gatherers Today, M. G. Bicchieri, ed., pp. 327-68. New York. 1968 What Hunters Do for a Living; or, How to Make out on Scarce Resources. In Man the Hunter, R. B. Lee and I. DeVore, eds., pp. 30-48. Chicago: AIdine. Lee, R. B., and I. DeVore 1968 Problems in the Study of Hunters and Gatherers. In Man the Hunter, R. B. Lee and I. DeVore, eds., pp. 3-20. Chicago: AIdine. Lumley, H. de 1972 A Paleolithic Camp at Nice. Scientific American 220:42-50. Lyon, P. ]. (ed.) 1974 Native South Americans. Boston: Little, Brown. Marshall, L. The !Kung Bushmen of the Kalahari Desert. In Peoples ofAfrica, ]. L. 1965 Gibbs, Jr., ed. New York: Holt, Rinehart and Winston. Martin, P. S. 1967 Pleistocene Overkill. Natural History 76:32-38. Meyer-Rochow, V. B. 1973 Edible Insects in Three Different Ethnic Groups of Papua and New Guinea. American Journal of Clinical Nutrition 26:673-77. Moore, F. 1970 Food Habits in Non-Industrial Societies. In Dimensions of Nutrition, j. Dupont, ed., pp. 181-220. Boulder: Colorado Associated University Press. Mountford, C. P. 1974 Walkabout. (A documentary film on the Australian Aborigines in the early 1940s.) Sydney: Film Australia. Murdock, G. P. 1934 Our Primitive Contemporaries. New York: Macmillan. 1959 Africa: Its People and Their Cultural History. New York: McGraw-Hill. Myers, N. 1982 "Homo Insectivorous." Science Digest, May. Oliver, D. L. 1967 A Solomon Island Society. Boston: Beacon Press. Ottenberg, S., and P. Ottenberg (eds.) 1960 Cultures and Societies ofAfrica. New York: Random House. Owen, R. C.; ]. ]. F. Deetz; and A. D. Fisher (eds.) 1967 The North American Indians: A Sourcebook. New-York: Macmillan. Perry, R. 1976 Life in Forest and Jungle. New York: Taplinger. Peterson, F. 1959 Ancient Mexico. New York: G. P. Putnam.


8. The Preference for Animal Protein and Fat

Potts, R., and P. Shipman 1981 Cutmarks Made by Stone Tools on Bones from Olduvai Gorge. Nature 291:577-80. Pyne, S. J. Indian Fires. Natural History 92(2):7-10. 1983 Quin, P. ]. Foods and Feeding Habits ofthe Pedi. Johannesburg: Witwatersrand Uni1959 versity Press. Radcliffe-Brown, A. R. 1964 The Andaman Islanders. New York: Free Press. Remington, C. L. 1946 Insects as Food in Japan. Entomology News 57(5): 119-21. Rivers, W. H. R. 1906 The Todas. New York: Macmillan. Roscoe, ]. The Baganda. London: Macmillan. 1911 Rowe,]. H. Inca Culture at the Time of the Spanish Conquest. In Handbook ofSouth 1946 American Indians, Bulletin 143, vol. 2, J. H. Steward, ed., pp. 183-330. Bureau of American Ethnology. Washington, D. C.: Smithsonian Institution. Ruddle, K. 1973 The Human Use of Insects: Examples from the Yukpa. Biotropica 5(2): 94-101. Sahagun, B. de 1963 Earthly Things. In General History of the Things ofNew Spain, book 11. Sante Fe: School of American Research and University of Utah. Schaller, G. B., and G. Lowther 1969 The Relevance of Carnivore Behavior to the Study of Early Hominids. SouthwestJournal ofAnthropology 25:307-41. Service, E. R. 1958 A Profile ofPrimitive Culture. New York: Harper and Row. 1971 Profiles in Ethnology. New York: Harper and Row. 1978 Profiles in Ethnology. 3d ed. New York: Harper and Row. Silberbauer, G. 1981 Hunter/Gatherers of the Central Kalahari. In Omnivorous Primates: Gathering and Hunting in Human Evolution, R. S. O. Harding and G. Teleki, eds., pp. 455-98. New York: Columbia University Press. Solecki, R. S. 1971 Shanidar: The First Flower People. New York: Knopf. Spencer, R. F.; ]. D. Jennings; et a1. 1965 The Native Americans. New York: Harper and Row. Starin, E. D. 1981 Monkey's Moves. Natural History 90:37-43.


III. Nutritional and Biopsychological Constraints

Stefansson, V. 1957 The Fat of the Land. New York: Macmillan. 1960 Food and Food Habits in Alaska and Northern Canada. In Human Nutrition: Historic and Scientific, Iago Galston, ed., Institute of Social and Historical Medicine, Monograph no. 3. New York: International Universities Press. Steward, ]. H. 1963a Handbook ofSouth American Indians, Vol. 4. New York: Cooper Square Publishers. 1963b Handbook ofSouth American Indians. Vol. 6. New York: Cooper Square Publishers. Steward, ]. H., and L. C. Faron 1959 Native Peoples of South America. New York: McGraw-Hill. Steward, ]. H., and A. Metraux 1948 Tribes of the Peruvian and Ecuadorian Montana, In Handbook of South American Indians, vol. 3, ]. H. Steward, ed., pp. 535-656. Bureau of American Ethnology. Washington, D. C.: Smithsonian Institution. Strum, S. C. 1981 Process and Products of Change: Baboon Predatory Behavior at Gilgil, Kenya. In Omnivorous Primates: Gathering and Hunting in Human Evolution, R. S. O. Harding and G. 1eleki, eds., pp. 255-302. New York: Columbia University Press. Tannahill, R. Food in History. New York: Stein and Day. 1973 Taylor, R. L. Butterflies in My Stomach. Santa Barbara: Woodbridge. 1975 Teleki, G. 1973 The Predatory Behavior of Wild Chimpanzees. Cranbury, N.].: Bucknell University Press. 1981 The Omnivorous Diet and Eclectic Feeding Habits of Chimpanzees in Gombe National Park, Tanzania. In Omnivorous Prilnates: Gathering and Hunting in Human Evolution, R. S. O. Harding and G. Teleki, eds., pp. 303-43. New York: Columbia University Press. Tonkinson, R. 1978 The Mardudjara Aborigines. New York: Holt, Rinehart and Winston. Treistman, ]. M. 1972 The Prehistory of China. Garden City, N. Y.: Natural History Press. U. S. Department of Agriculture 1975 Handbook of the Nutritional Content of Foods. New York: Dover. Vaidyanathan, A. 1983 Vegetarianism and Animal Protein Consumption in India. Paper read to the Symposium on Food Preferences and Aversions, sponsored by the Wenner-Gren Foundation for Anthropological Research. Van Lawick-Goodall, ]. In the Shadow of Man. Boston: Houghton Mifflin. 1971


8. The Preference for Animal Protein and Fat

Warner, W. L.

1958 1961
Weyer, E., Jr.

A Black Civilization. New York: Harper and Row. Social Lzfe ofEarly Man. New York: Wenner-Gren Foundation.

Washburn, S. L.

1961 Primitive Peoples Today. New York: Doubleday. White, E.; D. Brown; and the editors of Time-Life Books. 1973 The First Men. Boston: Little, Brown. Williston, S. W. 1883a Dipterous Larvae from the Western Alkaline Lakes and Their Use as Human Food, I. In Transactions (Connecticut Academy of Arts and Sciences) 4: 83-86. 1883b Dipterous Larvae from the Western Alkaline Lakes and Their Use as Human Food, II. In Transactions (Connecticut Academy of Arts and Sciences) 4: 87-90. Wittfogel, K. A. 1960 Food and Society in China and India. In Human Nutrition: Historic and Scientific, I. Galston, ed. Institute of Social and Historical Medicine, Monograph no. 3. New York: International Universities Press. Zihlman, A. Dietary Divergence of Apes and Early Hominids. Paper read to the 1983 Symposium on Food Preferences and Aversions, sponsored by the Wenner-Gren Foundation for Anthropological Research.



Biocultural Consequences of Animals Versus Plants as Sources of Fats, Proteins, and Other Nutrients
THIS CHAPTER DESCRIBES THE PHYSIOLOGICAL AND SOCIOcultural consequences of widely divergent diets based on either high animal or high plant intake. I examine variability in nutrient needs and nutrient sources and the benefits and liabilities of deriving nutrients from either animals or plants, using an evolutionary-adaptational framework to discuss population variability and the biochemical consequences of over- or under-consumption of protein, fats, and other nutrients.

Human Nutrient Needs
The long-established broad classification of the functions of nutrients in the body is still valid. Nutrients function (1) to supply energy; (2) to promote the growth and repair of body tissues; and (3) to regulate body processes. The nutrients that perform these functions are divided into five main categories: carbohydrates, lipids, proteins, minerals, and vitamins. A classification of the essential nutrients in each of these broad groupings is presented in Figure 9.1 (Guthrie 1975). There is one essential carbohydrate: glucose; one essential lipid: linoleic acid; nine essential amino acids; 15 essential minerals; and 13 essential vitamins. In addition, both water and fiber are needed. Sources of energy are provided by carbohydrates, lipids, and proteins, but minerals and vitamins are also necessary to catalyze the use of these nutrients. The nutrients that humans need have been well established. However, the quantities needed of these different nutrients have been under revision for a number of years and are still an area of controversy. The estimates of nutrient


III. Nutritional and Biopsychological Constraints

FIGURE 9.1. Essential Nutrients for Humans
Water Fiber Carbohydrate

Fat or Lipid
Linoleic acid

Protein (Amino Acids)
Histidine Isoleucine Leucine Lysine Methionine Phenylalanine Threonine Tryptophan Valine

Calcium Chlorine Chromium Copper Fluorine Iodine Iron Magnesium Manganese Molybdenum Phosphorus Potassium Selenium Sodium Zinc



Pyridoxine (B(-») Riboflavin (B L ) Thiamine (B l ) Vitamin (B IL )


Ascorbic acid (C) Biotin Folacin Niacin Panthothenic acid SOURCE: Guthrie (1975: 10-11).


9. Biocultural Consequences of Animals Versus Plants

requirements are determined by a number of techniques: (1) collection of data on nutrient intake from the food supply of normal, healthy people; (2) review of epidemiological observations when clinical consequences of nutrient deficiencies are found; (3) biochemical measurements that assess degree of tissue saturation and molecular function; (4) nutrient balance studies that measure nutritional status; (5) studies of subjects maintained on diets containing marginally low or deficient levels of nutrients followed by corrective measures; and (6) extrapolation from animal experiments in which deficiencies are not corrected. The results of these studies indicate that there is tremendous individual and population variability in all the biochemical processes involved in processing food. The biologist Rene Dubos has stated that "every person has a nutritionalmetabolic pattern that is characteristic, indeed as unique as a fingerprint" (1980: 14). One classic study of inter-population differences evaluated the protein requirements of American Caucasian males and Taiwanese Oriental males. Urinary nitrogen loss was significantly greater among the American students (Scrimshaw and Young 1978). In other studies comparing individuals, protein utilization differences of over 100 percent were observed for some essential amino acids. The reasons for variation in protein needs might include the following: (1) differences in absorption capacity; (2) differences in urinary nitrogen excretory rates; (3) differences in hormonal response to anxiety or stress; (4) differences in sweat loss during physical acdtvity; (5) differences in energy or carbohydrate intake and need; and (6) unknown genetic and environmental variables (Scrimshaw and Young 1978). (See Chapter 6 for further discussion of biochemical variation.) Because of these and other experimental findings, in the early 1970s the commissions on nutrition of the World Health Organization and the Food and Agricultural Organization scaled down the protein requirements of healthy persons some 30 percent (FAO 1973). As Dubos (1980) points out, this eliminated, by administrative decision, the international "protein gap" that had been postulated in earlier decades. It did not, however, eliminate the controversy over protein needs.

The Quantification of Nutrients (Recommended Dietary Intakes)
A brief discussion of the U.S. recommended dietary allowances (RDAs) for proteins and amino acids will serve to illustrate the points of controversy concerning the establishment of appropriate nutrient intake levels. The RDAs prescribe daily amounts of essential nutrients for population groups described


III. Nutritional and Biopsychological Constraints

by age, sex, weight, and physiological state (i. e., whether pregnant or lactating or not) (Committee on Dietary Allowances 1980). Special conditions that are not considered by those who establish the RDAs include body stores, physical activity, climate, and clinical or pathological problems of individuals Oackson 1983; Rennie and Harrison 1984). Nine amino acids-histidine, isoleucine, leucine, lysine, methionine, phenylalanine, threonine, tryptophan, and valine-are not synthesized or (in the case of histidine) are synthesized in inadequate amounts. They are, therefore, indispensable dietary nutrients for humans: the essential amino acids. Three steps were involved in establishing the RDA for protein: (1) estimation of the minimum amount of good-quality protein needed for maintenance of nitrogen equilibrium in healthy individuals; (2) adjustment of the requirements to allow for poor utilization of proteins from a mixed diet of good- and poor-quality protein (i. e., primarily plant foods with unbalanced amino acid ratios); (3) adjustments in protein allowance to meet the added needs of growth, pregnancy, and lactation. The amino acid patterns for high-quality proteins in milligrams per gram of protein are presented in Table 9.1, with requirements for infants and adults. Requirements are high for leucine and for phenylalanine. However, the sulpha-containing amino acid methionine is the one most often in limited supply. Since a specific ratio of amino acids is needed to produce a specific protein, when one amino acid is in low concentration the other amino acids, although

TABLE 9.1. Estimated Amino Acid Requirements of Humans
Requirements (mg/kg body weight/day) Amino Acid
Histidine Isoleucine Leucine Lysine Methionine and cystine Phenylalanine and tyrosine Threonine Tryptophan Valine


«4 yrs)
33 83 135 99 49 141


Amino Acid Pattern in High-Quality Protein (mg/g of protein)


42 70 51 26 73 35 11 48

21 92

16 12 10 16 8 3 14

SOURCES: Committee on Dietary Allowances (1980:43); Nutrition Search, Inc. (1975:222).


9. Biocultural Consequences of Animals Versus Plants

TABLE 9.2. Average Percentagea of Methionine Supplied in an Average Serving of Foods by Food Group
No. of Foods in Each Group
52 30 18 17 21 14 51

Food group
Breads, cereals, grains, and grain products Dairy products Fish and seafood Fruits Meats and poultry

Mean % of Methionine
10.6 30.0 174.8 2.7 130.1 77.6 22.6

Highest Source of Methionine
Soy flour: 84% 1cup Cottage cheese: 166%/cup Flounder: 268%/8 oz. Dried dates: 8%/cup Leg of lamb: 196%/8 oz. Sesame seeds: 192% I cup Soybeans: 46% 1cup

and seeds


SOURCE: Nutrition Search, Inc. (1975: 223-29). apercentage of RDA for an adult male weighing 160 pounds.

abundant, will not be used in the production of proteins. Table 9.2 lists the mean percentage of methionine supplied in an average serving of food from each of the seven major food groups. Fish, seafood, red meat, and poultry provide over 100 percent of the RDA of this nutrient. These foods provide approximately six to eight times the methionine found in an average serving of vegetables. Fruits, breads, and cereal grains provide very little methionine. The protein allowance that serves as the basis for the recommendations of the U.S. National Academy of Sciences/National Research Council (NASI NRC) is the 0.57 grams of protein per kilogram of body weight per day recommended by the Food and Agricultural Organization/World Health Organization (1973). But the NAS/NRC's Committee on Dietary Allowances has raised this figure 30 percent to take account of the lower efficiency of different types of proteins compared with the reference protein, which is that derived from eggs (Table 9.3). In addition, there is a correction of 7 percent to allow for variations in the efficiency of protein utilization in mixed diets. These adjustments lead to the RDA of 0.80 grams per kilogram of body weight per day. This is 56 grams of protein for a 70-kilogram man and 44 grams of protein for a 55-kilogram woman. This pattern of tinkering with established dietary allowances has a long history. (See Chapter 6 for an update of the FAa/WHO position. ) . In this connection it is instructive to review the protein requirements for one-year-olds issued by the NAS/NRC. In 1948 they established the need as


III. Nutritional and Biopsychological Constraints

TABLE 9.3. Protein Quality
Amino Acid Score a
100 66-70 44 41 47 28-43

Protein Source
Egg Meat, fish, poultry Whole-wheat flour Corn Soybeans Legumes

Biological Value b
94 74-76 65 60 71 55

Net Protein Utilization c
94 57-80 40 51 61 48

Protein Efficiency Ratio d
3.92 2.30-3.55 1.53 1.12 2.32 1.65

SOURCE: After FAa (1970, 1973). mg. of ami~o aci? i~ 1 gm. test prot~in x 100 mg. of ammo aCId m reference proteIn b dietary nitr~gen - ~urinary nitrogen.+ fecal nitrogen) x 100 dIetary nItrogen - fecal nItrogen c nItrogen retained nitrogen intake d weight gain . x number of weeks gm. pro t em

3.30 grams of protein per kilogram of body weight. In 1964 this was reduced to 2.50 grams, in 1968 to 1.80 grams, and in 1974 to 1.35 grams; finally, in 1980, it was raised again to 2.00 grams (Committee on Dietary Allowances 1980; Waterlow and Payne 1975). In general, the RDAs for other nutrients have not fluctuated as widely as those for proteins, although the establishment of RDAs has been more difficult for some nutrients than for others. Nevertheless, all of the R DAs call for overabundant quantities of nutrients rather than marginal levels. One rationale for maintaining high RDAs is to take into account the loss of nutrients through such processes as cooking, freezing, refining, and fermenting foods. The essential fat linoleic acid may be lost through high-temperature cooking, and a number of other vitamins are also heat-labile. Refining of grains, in particular, leads to a loss of fiber, vitamins, and minerals. Table 9.4 displays the reduction in specific nutrients through common processing techniques. In spite of these nutrient losses, our nutrient needs, even at the generous levels of the RDAs, are generally met by a "well-balanced" diet. But what happens when diets are not "well-balanced," when there is not a nutritionally appropriate mix of plant and animal foods?


9. Biocultural Consequences of Animals Versus Plants

TABLE 9.4. Reduction in Nutrient Content Through Processing
Fiber Fat (linoleic acid) Vitamin E Vitamin B 1 Vitamin B 2 Vitamin B6 Folic acid Vitamin B 12 Pantothenic acid Biotin Vitamin C Magnesium Zinc Iodine



R efininglOther

X X X Xa X X X X X X X X X X X X

SOURCE: Harris and Karmis (1975); Wing and Brown (1979:62). NOTE: Nutrients not listed are usually stable under these conditions. a Exposure to acid or alkaline pH.

The Biocultural Consequences of Diets High in Animal Foods
The heterogeneity of human diets reflects our omnivorous preferences and our biochemical flexibility. The shapes and functions of our digestive organs, our enzymes, and those of our resident intestinal bacteria enable us to extract the necessary nutrients from a wide variety of foods. Harris (1979) and others (see Chapter 8) have noted that in spite of our alimentary flexibility, there is a panhuman preference for meat and other animal products (i. e., milk, eggs, blood). This meat-eating preference, which is evident in the earliest hominid records, may be explained, in part, by the metabolic efficiency of meat as a source of human nutrition (Stini 1971). Meat provides amino acids in the appropriate proportions; essential minerals (e. g., iron and zinc); vitamins that are often difficult to obtain elsewhere (e. g., B 12 ); fat (although meat is relatively low in the essential fat linoleic acid); and some glucose, especially in raw mus-


III. Nutritional and Biopsychological Constraints

cle meat and liver. Moreover, meat provides these nutrients in an easily digested form and is nutrient-dense. That is, meat is a concentrated form of macro- and micro-nutrients; and therefore much smaller quantities of meat and other animal foods than of vegetables are needed to fulfill the RDAs for protein, fat, iron, zinc, and other nutrients. For example, to obtain eight grams of protein you could eat as little as one ounce of cooked lean meat or one extra large egg but you would need to eat four slices of enriched bread or four medium-sized potatoes for the same protein intake. Finally, the preference for meat and animal foods may involve specific organoleptic properties that appeal to our senses (see Chapter 7). Although our selections are culturally mediated, a number of traits transcend cultural differences. Humans, other primates, and rats are capable of learning to accept or reject foods on the basis of metabolic consequences. Primates are suspicious of new foods, but they also wish to explore and sample them (E. Rozin and P. Rozin 1981). Humans have a strong taste preference for sweet items. This preference is evident prenatally and is reinforced by the sweet taste of mothers' milk, which is high in lactose. The preference for sweet substances shows cross-cultural (Moskowitz et al. 1975) and age differences (Desor, Greene, and Miller 1975). Younger humans prefer higher concentrations of sucrose and lactose than do adults. They also prefer greater saltiness (Desor, Greene, and Miller 1975). The Rozins (1981) suggest that the preference for sweet tastes and the aversion to bitter or irritant tastes may be an important pressure in the course of human evolution. Potential human toxins often have a bitter taste. Other specific organoleptic preferences in humans involve food texture, temperature, and smell. Rosenbaum (1976) focuses on "mouthfeel." He suggests that humans like foods that are crispy, chewy, crunchy, and smooth. These are foods that can be controlled. We, in general, find unpleasant foods that are gummy, rubbery, stringy, slimy, mushy, or slick. Some of these textures require a great deal of mastication, while others may give the individual a feeling of being out of control as they slide down the throat. Food preparation techniques alter textures. Meat may be rubbery and stringy, but it is rarely slimy or mushy, and cooking may render meat more desirably chewy, crispy, or even crunchy.

Nutrients from Animal Foods
Diets that are high in animal foods usually provide abundant protein, fat, and calories (see Chapter 8). The American diet is high in both calories and nutrients derived from animal products. The "average" American family of four annually consumes 400 pounds of beef, 229 pounds of other red meat, 260 pounds of poultry, 1,233 pounds of dairy products, 1,056 eggs, 81 pounds of


9. Biocultural Consequences of Animals Versus Plants

seafood, 336 pounds of fresh fruit, 400 pounds of fresh vegetables, and 498 pounds of flour and other cereal products (Gainesville Sun, June 1982). Foods of animal origin provide most of the 75 grams of protein comsumed daily by the "average" American, but account for a mere 12 percent of the daily caloric intake. At the low end of the scale, estimated intakes of protein range from 10 to 40 grams per day for populations in Papua, New Guinea (Robson and Wadsworth 1977) to 200 grams of protein per day for some Inuit males (Draper 1977; Dubos 1980; Schaefer 1981). For the traditional hunting Inuit, protein supplies approximately 32 percent of the daily caloric intake. Fat intake is also very high, averaging 185 grams per day and representing 66 percent of the daily caloric intake (Draper 1977). In contrast, the first National Health and Nutrition Examination Survey indicates that U.S. males have an average daily intake of 100 grams of fat, and females 67 grams. For both groups fat accounts for 38 percent of the daily caloric intake (NCHS 1979). Other studies cite higher fat intakes, up to an average of 117 grams per day for adults (Draper 1977). Males consume more meat than females; and, as a consequence, a high percentage of their fat intake comes from meat. Children and adolescents of both sexes consume more dairy products than adults, and a larger proportion of their total fat and saturated fat comes from these animal sources. Eggs, meat, and dairy products contribute 73 percent and 67 percent of the cholesterol intake in males and females, respectively, in all age groups. Children and adults get the largest percentage of cholesterol from eggs, whereas adolescents report more cholesterol intake from meat. Linoleic acids are provided by fats and oils (36 percent) and poultry, particularly fried chicken (12 percent), for Americans of all ages. Meat is the major source of oleic acid (another important unsaturated fatty acid) for adults, and milk and milk products the major source for children (NCHS 1979). Lipids or fats provide the most concentrated source of dietary energy, containing twice the number of calories per gram as protein or carbohydrates. Fats are usually obtained from animal sources and oils from plant sources. As lipidprotein complexes, they are found as components of cell membranes and as components of steroid hormones. Fat serves as a fuel when glycogen stores are depleted. Subcutaneous layers of fat are energy reservoirs and insulate the body against cold. Fat also serves to insulate vital organs. Recent evidence indicates that there are individual and possibly population differences in the amount and activity levels of thermogenic brown fat. These differences are under genetic control and account for substantial differences in basal metabolic rates and, hence, nutrient and caloric needs (Bray 1983). The high-energy-yielding diet of Inuits is necessary to maintain basal metabolic rates that are 13 to 33 percent higher than those of populations in temperate climates (Draper 1977; Speth and Spielman 1983). The task of keeping warm in cold climates is calorically costly as indicated by estimated daily caloric


III. Nutritional and Biopsychological Constraints

intakes of 3,600 kilocalories for adult males (Draper 1977). These needs are about one-third greater than the United States RDA for energy. Furthermore, protein metabolism itself is energetically costly. Nearly one-third of all ingested protein calories are used to metabolize protein and in the process produce body heat. This specific dynamic action is about two to three times higher than one finds in a diet consisting largely of fat and about five times higher than one finds in diets high in carbohydrates (Speth and Spielman 1983). Additionally, proteins, once digested, may be used in the production of carbohydrate energy sources when carbohydrate intake is very low. Again the traditional Inuit diet serves as an example, with an estimated intake of 10 grams of carbohydrates per day accounting for only 2 percent of the daily caloric intake (Draper 1977). Although fatty acids may be used for most energy needs, glucose is the metabolically preferred source of energy, and some organs, such as the brain, are obligate glucose-users. Not surprisingly, the Inuit are extremely efficient at the production of glucose from amino acids: that is, the process of gluconeogenesis (Draper 1977; Lieberman n. d.). This metabolic adaptation to a long-term nutrient imbalance represents a genototrophic adaptation among Inuit and Aleuts. Unfortunately, it becomes maladaptive when carbohydrate intake rapidly increases as part of other dietary and lifestyle changes (Schaefer 1971; Weiss, Ferrell, and Hanis 1984). As Schaefer (1971) notes, for one group of urbanized Canadian Eskimos, 50 percent of the total caloric intake came from carbohydrates. There was a corresponding decrease in animal protein consumption. Although deficits in energy and nutrients usually occur when animal protein intake is low, some hunter-gatherers in northern latitudes who suffer deficiencies in spite of high meat intake are exceptions. Speth and Spielman (1983) describe a fall and winter diet high in lean meat but low in fat and carbohydrates for some Canadian and northern U. S. natives. Traditional food animals (e. g., deer, bison, moose, caribou) lose body fat as a result of the fall rut for males, reduced forage for both sexes in the winter, and pregnancy and lactation for females in the early spring. The whole-body fat content in wild ungulates ranges from a high of 15 to 30 percent to a low of 1 to 2 percent. The protein content of lean meat remains relatively constant at about 21 percent (Speth and Spielman 1983: 12). As noted above, diets high in protein have a high specific dynamic action that requires additional energy for metabolism. The estimated caloric increase is 20 percent above basal metabolic rates. For example, if the basal metabolic rate required 1,600 kilocalories, an additional 320 kilocalories would be needed for specific dynamic action (Speth and Spielman 1983: 5-7). These authors estimate that to meet daily energy needs of approximately 3,600 kilocalories, a hunter would need to consume 7.5 to 7.9 pounds of lean meat. When caloric needs are not met, the body uses available stores of fat and then skeletal muscle protein for energy. Both fat and carbohydrates have a protein-sparing


9. Biocultural Consequences of Animals Versus Plants

effect, but because glucose is a required energy source for the central nervous system, carbohydrates provide a more efficient and protein-protective source (Speth and Spielman 1983: 13-15). A diet high in lean meat without additional carbohydrate and fat sources presents the possibility of linoleic acid deficiency. Meat is only a fair source of this essential fatty acid. Better sources are oil-rich seeds and grains. Low fat intake results not only in reduced energy sources, but also in a reduction of lipoprotein production and the absorption, transportation, metabolism, and storage of fat-soluble vitamins (Reed 1980). Furthermore, diets high in animal protein inhibit the absorption of calcium. Even Americans have low intakes of calcium, and these low intakes, perhaps coupled with a reduction in absorption ability, have been implicated in both osteoporosis and high blood pressure (Tufts University Newsletter 1984b; McCarron et al. 1984). Speth and Spielman (1983) suggest a number of specific strategies employed by hunter-gatherers to offset the dietary deficiencies of high-protein, low-carbohydrate, low-fat diets. For example, there may be selective hunting of individuals or species with higher body fat or the consumption of fat-rich portions of the animal. Labor-intensive techniques may be employed to render fat from bones or to preserve fat. There may be selective gathering of oil-rich foods and the trading of animal items for high-fat or high-carbohydrate foods. Finally, humans may consciously gorge on fatty foods when they are available so that their fat stores may be drawn on in times of energy deficiencies (Wilmsen 1978). This latter strategy has been useful in populations confronted with the unusual circumstances described above, but in sedentary, Westernized populations the periodic or chronic overconsumption of protein and fat have been linked to a number of disorders.

Animal Foods: Dangers and Disorders
There are many dangers inherent in keeping and consuming animals and animal products. Herd animals may transmit diseases to humans (e.g., anthrax and brucellosis) or parasites such as Trichinella spiralis. Animal muscle and fat may be contaminated by plant substances ingested by the animal; by antibiotics, hormones, and other drugs in animal feed; by pesticides and herbicides; by radioactive substances; by naturally occurring fungi and bacteria producing human toxins; and by the addition of potential carcinogens during preparation and cooking (Liener 1974). In the United States both diary products and eggs have been contaminated with pesticides and radioactive substances, and milk has been contaminated with pathogenic bacteria and fungal mycotoxins (ibid.). Furthermore, nutrients and their metabolites may prove to be pathogenic in humans if consumed in large amounts. The classic example is hypervitaminosis


III. Nutritional and Biopsychological Constraints

caused by the high levels of consumption of the livers of marine mammals, known as "polar bear liver disease." Humans are not infrequently allergic to milk and egg proteins and metabolically intolerant of lactose, the primary sugar in milk (Harrison 1975; McCracken 1971; Simoons 1969, 1970, 1980). Lactose intolerance results from the reduction in older children and adults of the production of the enzyme lactase. Lactase splits the disaccharide lactose into the component sugars galactose and glucose. Some individuals with lactose intolerance can consume other dairy products without gastric and intestinal distress because bacteria reduce the lactose content of cheese, yoghurt, ghee, and other dairy products. Anthropologists are interested in lactose intolerance because its frequency varies so widely in human populations. Lactose tolerance or lactase sufficiency, unlike other diet-related polymorphisms, shows marked geographical differences. These differences are specifically related to a history of milk consumption and dairying. Among populations with high milk production and high consumption by adults, there is very low frequency of lactase deficiency (approximately 15 percent). Among groups with no or low consumption of milk, over 80 percent of the population may be lactase-deficient (Harrison 1975; McCracken 1971). The advantage of consuming milk is that it enhances the absorption of both calcium and protein when lactose can be digested. Harrison (1975) argues that this enhanced calcium absorption has been a particularly powerful selective force under environmental conditions with limited sunlight and limited sources of vitamin D. The utilization of milk would have an advantage in reducing the risk of rickets, which is known to directly affect the pelvic as well as the lower limbs, enhancing both infant and maternal mortality. Calcium is also needed for various neurological functions whose impairment may compromise fertility. Harrison (1975) has indicated that a very small selective advantage of 1 percent for individuals carrying the gene for lactase production in adulthood would have caused the prevalence of lactose intolerance to decline from 90 percent to 16 percent of the population in 400 generations, or since the domestication of sheep and goats. This would mean an increase in the gene frequency for lactase production from 5 percent to 60 percent. Lactose comes only from animal sources, and therefore selection pressure is heaviest on those individuals who use milk and milk products as adults to obtain calcium and other nutrients. Although milk drinking has nutritional advantages, Simoons (1981) notes an increased risk of cataracts in milk-drinking populations. He suggests that this might be related to the high concentration of galactose in milk and milk products. Of all the diet-health issues, for decades the most conspicuous has been the possibility of a dietary influence on coronary heart disease (National Livestock and Meat Board 1983). The hypothesis most widely investigated is that saturated fats or dietary cholesterol or both are high-risk factors for heart disease.


9. Biocultural Consequences of Animals Versus Plants

A fairly consistent epidemiological finding has been a positive correlation between elevated blood cholesterol levels and chances of developing atherosclerosis. Consistent findings demonstrate that egg-, milk-, and meat-eaters are at higher risk for heart disease. These sources account for 73 percent of the daily intake of cholesterol and 57 percent of the saturated fat intake in the U.S. diet (NCHS 1979). The average daily U.S. diet contains about 500 milligrams of cholesterol for men and 350 milligrams for women (National Livestock and Meat Board 1983), in both cases about 60 percent more than is recommended by the American Heart Association. In actuality, humans are poor digestors of cholesterol, and as little as 10 to 50 percent of it is absorbed. The body can manufacture from 800 to 1,500 milligrams of cholesterol daily even if no cholesterol is consumed in the diet. A national long-term study that followed men with elevated cholesterol levels for a period of 7 to 10 years found that those on a moderately cholesterollowering diet plus blood-cholesterol-lowering drugs reduced their cholesterol levels by 8.5 percent (Tufts University Newsletter 1984a:3-5). The researchers concluded that for each 1 percent reduction in cholesterol there was a 2 percent reduction in the rate of heart attacks. However, other long-term studies have not shown these results (MRFITRG 1982). Rosch concluded in a letter to the Lancet: "It seems quite clear that reduction of fat, in contrast to caloric restriction, has little to do with blood cholesterol. In addition, the incidence of heart attacks is lowest in midwest U.S.A. farming areas, where fat consumption is the highest" (1983:851). He suggests that the results of studies that have shown reduction in heart attacks may be due to the effective lowering of stress-related catecholamines. These honnones can cause cardiac damage (Rosch 1983). The stress-reducing value systems and lifestyles found among some vegetarian groups may be partially responsible for the finding that vegetarians have not only lower overall cholesterol levels but a higher percentage of protective high-density lipoproteins than meat-eaters (Liebman 1983; Sims 1978). One large-scale study of 25,000 Seventh-Day Adventists over a 20-year period shows that lacto-ovo-vegetarians were much less likely to die of heart attacks than meat-eaters. Within this group of Seventh-Day Adventists, meat-eaters were twice as likely to die of heart disease as vegetarians (Snowdon, Phillips, and Fraser 1984; West and Hayes 1968). Their lower plasma cholesterol concentration and rate of heart attacks are most likely due to a number of factors, including: (1) lower caloric consumption; (2) lower consumption of saturated fats and higher consumption of polyunsaturated fats; (3) higher consumption of fiber; (4) lower salt intake; (5) lower body weight; and (6) lifestyle factors such as more exercise and less stress (Flynn 1978; Krey 1982). The above factors are also causally related to the prevalence of hypertension among populations of meat-eaters and non-meat-eaters. For example, there is an impressive positive relationship between salt intake and the incidence of


III. Nutritional and Biopsychological Constraints

hypertension across populations (Luft, Weinberger, and Grim 1982). Most populations exhibiting a high frequency of hypertension have a diet high in animal food (Gleibermann 1973). Meat contains more sodium (1 g of salt == 400 mg of sodium) than most plants, but, paradoxically, Americans get 24 percent of their sodium from grain products and only 9 percent from meat. The reason is that grain products are processed with salt as well as other sodium-based additives (NCHS 1979). Gleibermann (1973) has linked the history of salt consumption and nutritional behaviors to the high incidence of hypertension among New World black populations. She argues that the high prevalence of hypertension in blacks with a moderate salt intake is related to past selection pressures in tropical Africa. She hypothesizes a tremendous selection pressure to develop salt retention mechanisms in populations living in areas with high temperatures and low salt availability. Dietary salt intake would have decreased as populations moved from a gathering-hunting economy to horticulture. Although salt is a very early trade item, Gleibermann notes that individuals who were most likely to be stressed through heat exhaustion were less likely to participate in the exchange and consumption of salt. Under these circumstances, therefore, there would be a selective advantage to conserving salt. Salt retention is mediated through the renin-aldosterone system. Recent research has shown consistently low renin activity levels among black Americans. The result is that blacks excrete less sodium during the day and more at night than white subjects, and their 24-hour sodium excretion is significantly less than that of whites (Luft, Weinberger, and Grim 1982). The increased fluid volume leads to increased blood pressure. Although there is no specific genetic model for hypertension as there is for lactase deficiency, one interpretation of the link between hypertension and dietary sodium indicates that there may be a genetic factor in the absorption and retention of this mineral. If sodium retention is a phenotype that has been selected for, then hypertension may be seen as the disgenic outcome of a oncefavorable physiological adaptation. Increased salt sensitivity becomes disgenic in environments with moderate and frequent salt intake and low excretory rate through sweat and urine. Hypertension fits my model of a hyperefficiency disease (Lieberman n. d.). Recent studies using the large U. S. data base of over ten thousand individuals compiled in the first National Health and Nutrition Examination Survey have shown that hypertension is consistently associated with low calcium intake. Individuals with high blood pressure also have diets low in linoleic acid, carbohydrates, calories, potassium, and vitamins A and C. Surprisingly, lower mean systolic blood pressure and lower absolute risk for hypertension were associated with higher sodium intake (McCarron et al. 1984). When intakes were controlled for age, race, sex, and body mass index (weight in kilo-


9. Biocultural Consequences of Animals Versus Plants

grams/height in meters squared), intakes of calcium, potassium, and vitamins A and C were still found to be low in hypertensive individuals. The food group most highly correlated with blood pressure was dairy products. The greater the individual's consumption of dairy products, the less likely he or she was to be hypertensive (McCarron et al. 1984: 1396). Low sodium intake is also associated with lower intakes of dairy products. In this study, individuals reporting low sodium intakes were at two to three times greater risk for hypertension than those reporting high intakes. These authors suggest a number of biochemical and physiological mechanisms by which low levels of these nutrients could cause elevated blood pressures. Finally, obesity, as assessed by weight or body mass index, was positively correlated with hypertension but negatively correlated with caloric and other nutrient intakes. McCarron et al. (1984) conclude that nutritional deficiencies, and not excesses, distinguish overweight or hypertensive individuals in the United States. They warn that standard weight reduction diets may actually exacerbate conditions leading to hypertension by further reducing nutrients that are essential for maintaining normal blood pressures. Both excesses and deficiencies have been implicated in the etiology of New World syndrome, found among Native Americans and Mexican Americans (Weiss, Ferrell and Hanis 1984). The syndrome includes early-age adult obesity, type II (adult-onset, non-insulin-dependent) diabetes mellitus, the formation of cholesterol gallstones, and gallbladder cancer. Cancers at other sites and the complications of these primary diseases also occur in elevated frequencies. Weiss, Ferrell, and Hanis (1984) refer to this syndrome as an epidemic that has been dramatically increasing since World War II. They have formulated a thrifty genotype hypothesis to account for it because, like other "diseases of Westernization," it results from an interaction of susceptible genotypes and rapidly changing environments, including significant dietary changes. The now unfavorable metabolic phenotypes were selected for under the earlier conditions of northern hunter-gatherers. The relevant observations indicate that obesity is associated with New World syndrome; that the syndrome involves aspects of nutrient absorption, deposition, metabolism, and storage; that it is exacerbated at puberty and during pregnancy; and that it involves lipid physiology and hepatic function (Weiss, Ferrell, and Hanis 1984: 171). The authors suggest that genes related to the storage of food in the form of fat might have conferred a selective advantage on women during times of nutritional and physiological stress. This genotype was previously suggested as the mechanism by which diabetes manifested its thriftiness (Neel 1962). However, obesity alone does not account for the other components of the syndrome, and therefore the authors suggest that there may be another metabolic abnormality that predisposes the system to both obesity and increases in lithogenic bile leading to gallbladder stones.


III. Nutritional and Biopsychological Constraints

The diets implicated in the etiology of New World syndrome are high in fats, particularly saturated fats; low in fiber; high in carbohydrates. The first two components are characteristic of higWy animal-based diets, whereas the third is characteristic of both contemporary and traditional diets. However, although both types of diets are high in carbohydrates, traditional diets are also high in dietary fiber and often low in fat and total calories. The biocultural consequences of diets either high or low in fat have been discussed above. The low fiber content of meat-based diets has also received a lot of attention. A diet low in fiber increases the amount of time food stays in the intestines. Although this allows more time for the absorption of nutrients, many scientists have argued that enhanced absorption and increased transit time can lead to obesity, diabetes, atherosclerosis, colon cancer, and other diseases of "civilization" (Anderson and Ward 1978; Trowell and Burkitt 1981). The relationship between specific dietary components and cardiovascular and metabolic diseases is not always clear. The relationship of diet to a second disease cluster, cancer, is even more tenuous. Many cancers have been linked to specific dietary components (e. g., fat) consumed in high amounts in animalbased diets. Other cancers have been linked to food preparation techniques and food additives. Cancer is second to heart disease as a cause of death in the United States. Yet the epidemiological data present problems. For example, the U.S. gastric cancer rate of 9.2 per 100,000 deaths has declined more than 80 percent since 1930. In the same period of time, the U.S. per capita consumption of processed meat products has more than doubled. This means that an increase in nitrates and nitrites, food colorings, and additives has been accompanied by a reduction in cancers of the digestive tract organs. It is important to note that both animal products and plants have been cited as potentially carcinogenic. For example, we are urged to decrease our consumption of barbecued foods because of the benzopyrine that is formed when animal fat is burned at high temperature. On the other hand, the high nitrate content in vegetables such as spinach may be converted in the body to nitrite, react with the amino acids, and produce nitrosamines, which are known carcinogens. Common green vegetables contain as much as 2,000 parts per million of nitrate or more. Cured meats customarily contain 100 parts per million or less of nitrite and nitrate combined. Like nitrites and nitrates, nitrosamines are also widely distributed in the environment (National Livestock and Meat Board 1983). Even heating proteins at high temperatures may result in amino acid reactions leading to carcinogens. The essential amino acid tryptophan is especially susceptible to this reaction. Unsaturated fats, since they oxidize and form free radicals, can destroy or change RN A and DNA. Vegetables high in natural estrogens, like brussels sprouts and cabbage, may be considered potential inducers of genetic and carcinogenic changes (Alcantara and Speckman 1976).


9. Biocultural Consequences of Animals Versus Plants

Alcantara and Speckman (1976) review the literature on diet and nutrition and cancer. They state that diet and nutrition are more appropriately seen as modifiers rather than initiators of tumorigenesis. In addition to dietary factors, exposure to environmental carcinogens, genetic and immunological background, stress, the aging process, and obesity may be etiologic variables for specific cancers. These findings tend to bring into question the relationship between certain dietary components and specific cancers (National Livestock and Meat Board 1983). However, Phillips and Snowdon's work on Seventh-Day Adventists as reported by Liebman (1983) shows that Seventh-Day Adventist vegetarians are only about one-half as likely as other Americans to develop colon and rectal cancer. There is also a lower incidence of breast, prostate, pancreatic, and ovarian cancer. Phillips and Snowdon note that vegetarians not only eat less or no meat but also eat more beans, whole grains, vegetables, and other fruits. The fiber and vitamin A, C, and E content of their diet may be important in terms of preventing the developing of specific cancers. The findings on both cardiovascular diseases and cancer reveal that an overintake of calories leading to obesity, rather than an excess or deficit of any specific nutrient, from either an animal or a plant source, may be the key risk factor in these diseases. The disadvantage that animal-eaters have is that they are generally fatter than vegetarians. The relationship between leanness per se, dietary constituents, nutrient status, and diseases may be further investigated through an examination of diets high in plant foods.

Biocultural Consequences of Diets High in Plant Foods
Most of the world's populations live on diets characterized by very high intakes of plant foods and based on a starchy, bland staple (Robson and Wadsworth 1977). Most diets contain both plant and animal products. However, some traditionally vegetarian groups such as Orthodox Hindus, Buddhists, and SeventhDay Adventists, and a few smaller religious orders such as Trappist monks, maintain exclusively vegetarian diets. Moreover, the last two decades have seen the emergence of what Abrams (1980) and others have called the "new vegetarian." Whether traditional or new, vegetarianism is practiced in many forms. Strict vegetarians, often referred to as "vegans," eat diets that contain no animal products but consist of fruits, vegetables, seeds, nuts, and so on. Some vegans limit their diets to fruits, nuts, honey, and oil. More commonly, vegetarians will eliminate only meat, often red meat. There are a number of varieties: ova-vegetarians, who include eggs; polo-vegetarians, who include poultry; pesco-vegetarians, who include fish; lacto-vegetarians, who include


III. Nutritional and Biopsychological Constraints

TABLE 9.5. Sources of Nutrients Frequently Lacking in Vegan Diets
Riboflavin (B:J Vitamin B 12

Usual Sources
Dairy products, meat, eggs Dairy products, meat, eggs, fish

Alternative Sources
Brewer's yeast, leafy vegetables None (some foods may be fortified or contaminated with BITproducing bacteria)

Vitamin D Calcium

Fortified milk, fish, sun Dairy products, fish, leafy vegetables

Sun, fortified foods Soy products, leafy vegetables


Meat, eggs, leafy vegetables

Legumes, leafy vegetables, dried fruit, cookware


Meat, liver, eggs, seafood, cheese

Legumes, nuts, wheat germ, whole grains

SOURCE: After Liebman (1983: 11).

dairy products; and ovo-Iacto-vegetarians (Abrams 1980). As previously noted, vegans can achieve a balanced diet if they are careful about the consumption of specifically limited micro-nutrients such as B 12 , vitamin D, iron, and zinc (Table 9.5) and the balance of proteins (Bergan and Brown 1980; Committee on Nutrition 1977).

Nutrients from Plant Foods
Plant foods offer a variety of tastes, textures, colors, and aromas not found in animal foods. Vegetables and fruits and the condiments made from plants not only add nutrients to starchy staples (e. g., sorghum, wheat, corn, manioc, potatoes) but impart specific flavor characteristics (Rozin and Rozin 1981; see also Chapter 7). In addition, Rozin argues, certain foods may have specific desirable psychopharmacological effects. Common examples include coffee, for the stimulant effect of caffeine, and chili peppers, as possible inducers of endogenous opiate-like substances. Nutrient needs are generally met by a "well-balanced" vegetarian diet. Although plants, by volume or weight, are primarily sources of digestible and indigestible carbohydrates, they also supply protein, fat, and the micro-nutrients. It is possible to obtain enough protein, with appropriate amino acid


9. Biocultural Consequences of Animals Versus Plants

ratios, from a vegetarian diet, but it requires more than one plant food, which is not the case with meat. Plant proteins do not have the same composition as animal protein and are therefore deficient in some amino acids or have inappropriate amino acid ratios for the production of human proteins. If an amino acid is not available in the amino acid pool when it is needed for protein synthesis, the synthesis stops. The limiting amino acid in the human diet is usually methionine (see Table 9.3). Foods that are low in methionine are primarily of plant origin. The lowest percentage of methionine occurs in fruits, followed by breads and other grain products and by vegetables. There are also differences in the quality of proteins and the net protein utilization value (see Table 9.4). The lowest net protein efficiency ratios occur for corn, whole wheat flour, and legumes. The notion of complementary proteins, whereby one protein supplies the limiting amino acid not found in the other, is familiar to nutritionists and individuals in the health-food movement (Lappe 1975). Many traditional cuisines, such as those in Latin America, combine corn and beans to provide complete protein. The limiting amino acid in com is lysine, whereas the limiting amino acid in beans and other legumes is methionine. On the other hand, the combination of corn and wheat produces an incomplete protein because lysine is low in both com and wheat. Katz and his co-workers have investigated a number of food-processing techniques that enhance nutrient quality and quantity and reduce concentrations of toxic chemicals as a means of cultural accommodation to important dietary staples (Katz 1980; Katz, Hediger, and Valleroy 1975; Katz and Schall 1979; and see Chapter 5). Much of their work is relevant to the discussion of amino acid complementation (Katz, Hediger, and Valleroy 1975). Com (i.e., maize), a widely distributed New World staple crop, undergoes a processing technique in which it is soaked in an alkali or lime solution. The lime liberates the vitamin niacin from an indigestible complex and improves the amino acid quality of the digestible protein fraction. This treatment actually lowers the quantity of amino acids except lysine, in which corn is generally low, so that there is a relative enhancement of lysine in comparison with the other amino acids. Com also has a very high concentration of the amino acid leucine. An imbalance of the ratio of isoleucine to leucine has been implicated in the etiology of the niacin deficiency disease pellagra. In societies that traditionally treat com with an alkali, there is little or no pellagra present. Overall there is an enhancement of the amino acid ratio so that the bioavailability of the amino acids is increased. Com is a fairly good source of protein, which makes up 6.8 to 12.0 percent of the kernel weight. Soybeans are widely used in Asia. They are high in protein and calcium but contain a potent anti-trypsin factor that can only be deactivated by long exposure to high temperatures. Trypsin is an important digestive enzyme, found


III. Nutritional and Biopsychological Constraints

in pancreatic juices, that breaks down proteins into smaller polypeptides. A reduction in trypsin means a decreased efficiency of protein digestion. An investigation of the food preparation techniques employed in the manufacture of soybean products indicates that soy proteins (curd) were traditionally precipitated with magnesium in Japan to produce tofu and with calcium in China to produce doufu (Katz 1980). The curd is produced without the anti-trypsin factor. In the production of soy sauce, fermentation deactivates the anti-trypsin factor and enhances other nutrients, such as vitamin B 12 . Other traditional preparation processes include sprouting the beans and the production of soy milk. Soybeans are an important dietary constituent, contributing a major portion of protein in a vegetarian diet. Katz (1980) points out that in cuisines where soybean curd is used extensively, so is rice. These foods provide complementary amino acid compositions. Recent work by Sirtori and Lovati (1984) has demonstrated that the substitution of soybean protein for animal protein in a diet increases the favorable polyunsaturated/saturated fatty acid ratio and decreases plasma cholesterol levels in people with hypercholesterolemia. In their study the reductions in cholesterol levels always exceeded 15 percent and were independent of dietary cholesterol intake. The authors propose a mechanism by which soy proteins act on the liver receptors for lipoproteins. These receptors maintain normal or heightened affinities for lipoproteins and do not show the decreased activity level usually induced by excess dietary cholesterol. Other mechanisms operating on neutral steroids are also possible. These authors strongly endorse the use of soy protein diet therapy for people with hypercholesterolemia. Soy protein has little effect on people with normal cholesterol levels. Studies tend to indicate that circulating levels of cholesterol, the low-density lipoproteins that are associated with cardiovascular disease, and triglyceride levels tend to be lower for vegetarians than for non-vegetarians (Snowdon, Phillips, and Fraser 1984). Yet individuals on complete vegetarian diets (vegans) have little trouble obtaining appropriate dietary fats. As noted above, the essential dietary fat, linoleic acid, is found in greater concentrations in plant than in animal foods. Linoleic acid is a polyunsaturated fatty acid. Polyunsaturated fatty acids predominate in vegetable oils, whereas saturated fats occur primarily in animal fat. Once lipids are absorbed by the intestine, they undergo further breakdown by additional enzymatic action. These glycerides and fatty acids are then absorbed into the blood system for transport. The amount absorbed is about 50 percent of intake. Since proteins can be transported more readily than lipids in a watery medium, combinations of lipoproteins are formed. These lipoproteins are then transported to a number of sites, where they are active in a variety of metabolic processes. Lipids are an important energy source. They have twice the number of calories per gram of food as either proteins or fats, so that diets low in calories may be low in fats.


9. Biocultural Consequences of Animals Versus Plants

Carbohydrates fonn the third macro-nutrient class that is readily obtained from plant sources. There are a number of types of carbohydrates. Monosaccharides are often referred to as simple sugars. Glucose, the most abundant of these, is essential for energy metabolism. Most foods contain disaccharides. These are found in sugar (sucrose), fruits (fructose), milk (lactose), vegetables, and malt products. Glucose is the major simple sugar in the polysaccharides of seeds, grains, cereals, legumes, fruits, and vegetables. Free glucose (as a monosaccharide) is found in some fruits and vegetables, honey, and corn syrup. The polysaccharides include starch, which is found extensively in grains, cereals, legumes, root vegetables, and fruits. Carbohydrate digestion is quite efficient: about 98 percent of digestible carbohydrates in the diet are ultimately available for biological use. The chemical digestion of carbohydrates is accomplished primarily through specific enzymes in the mouth and small intestine. In general, the simpler sugars are soluble in water and impart sweetness. Fructose has the highest relative sweetness, followed by sucrose and glucose. Cellulose and starch lack this taste property. Other carbohydrates are indigestible for humans. The most common of these are cellulose and pectin, which serve as bulk in the human digestive tract. Cellulose is the major component of cell walls of plants and is ingested with all fruits and vegetables. Pectin is a jelly-like material that acts as cellular cement in plants. It is found abundantly in unripe fruit. Glycogen is sometimes referred to as "animal starch" because it is the principal carbohydrate storage material in animals. Foods with dietary fiber content greater than 20 percent of their dried weight include wheat bran, rye bread, endive, kale, and cabbage. Foods with low dietary fiber (less than 10 percent) include soybeans, peanuts, oats, rice, white bread, potatoes, and carrots (Guthrie 1975; Reed 1980). Dietary fibers decrease transit times, reducing the time for nutrient absorption. They also add bulk to the diet, so that infants and children on primarily low-processed vegetarian diets may have difficulty obtaining the RDAs for protein and other nutrients because of the high fiber content of the diet, their small digestive tracts, and their relatively higher nutrient needs. Carbohydrate consumption, primarily from vegetable sources, has been linked to a number of diseases in Western societies. Although these diseases have continued to increase in incidence, the actual consumption of carbohydrates in the United States, for example, has decreased approximately 25 percent since the turn of the century. Recent work by Jenkins (1982) and others has demonstrated that the distinction between simple and complex carbohydrates-that is, sugars and starches-is not meaningful metabolically. The concentration of blood glucose and the rise of the blood glucose curve after ingestion of specific carbohydrate loads are unrelated to this distinction. These findings have confounded the previous dietary prescriptions for the diabetic


III. Nutritional and Biopsychological Constraints

diet. For example, potatoes and bread produce very steep rises in blood glucose, whereas com and rice do not. The increases in blood glucose with potatoes and bread are equal to or exceed those caused by the ingestion of sugars. Furthermore, simple explanations based on the different fiber contents of foods and their different absorption times do not hold up under examination. Whole-grain bread and refined white bread, for example, have about the same blood glucose curve Oenkins 1982). Other physical properties of the molecules of starches and sugars may be important in the glucose response curve. It is clear that consuming carbohydrate foods with higher protein and fat content (e. g., ice cream, high in sugar and fat) decreases the blood glucose response. Both Type II diabetics and non-diabetics initially respond in the same way to these foods, but in individuals with diabetes the elevated glucose levels persist longer. Snowdon and Phillips (1985) have asked whether a vegetarian diet reduces the occurrence of diabetes. Since 1960 they have studied over 25,000 white Seventh-Day Adventists. Approximately 50 percent of the members of this religious group are lacto-ovo-vegetarians. They do not have the high rate of diabetes characteristic of Amerindians and black Americans and are generally at low risk for the disease. The risk of diabetes as an underlying cause of death for Seventh-Day Adventists is about half the risk for all U. S. whites. When comparing the two Adventist samples of meat-eaters and non-meat-eaters, the investigators found that vegetarians had a lower risk of dying with diabetes as an underlying cause and lower self-reports of diabetes. When the investigators controlled for the possible confounding factors of weight, physical activity, and other dietary factors (e. g., fat and sugar intake), the risk of diabetes remained lower for vegetarians. Interestingly, all the associations between meat eating and diabetes were stronger in males than females in this study (Snowdon and Phillips 1985). Six micro-nutrients have been shown to be more difficult to obtain from natural food sources if an individual has a diet exclusively composed of plant foods (see Table 9.6). In most cases supplements or fortified vegetable foods will provide the nutrient, as in the case of riboflavin (vitamin B2 ) and vitamin D. The nutrient that is most difficult to obtain on a vegan diet is vitamin B 12 . The natural sources of B 12 are dairy products, meats, eggs, and fish; there are no plant sources. Individuals on a vegan diet must get vitamin B 12 either through the contamination of the food by B l2 -producing bacteria that live in the soil or through fortified foods or dietary supplements. Iron deficiencies are common in both meat-eaters and vegetarians in the United States (Guthrie 1975; Reed 1980). The primary source of readily available iron comes from animals. This is the heme iron found in meat. It is absorbed more readily than iron of vegetable origin (Scrimshaw and Young 1978).


9. Biocultural Consequences of Animals Versus Plants

Some incompatible dietary combinations lead to a decrease in iron absorption. These include the commonly found phytates in whole grains and oxalates in spinach and other green leafy vegetables. These phytates and oxalates bind iron and other minerals such as zinc and calcium and make them unavailable for absorption. Absorption of iron can be enhanced when small amounts of red meat are added to the diet. The extent to which a diet is high in phytates and oxalates and low in heme iron from meat depends primarily on social, economic, and cultural factors. The RDAs based on age, sex, and physiological state do not take account of these other factors, which may contribute substantially to population and individual variation in iron needs. For example, enhanced sweating and high environmental temperatures increase the loss of iron from the body. Bacterial and viral disease as well as certain parasites such as hookworm, schistosomiasis, and malaria increase blood loss. In spite of these potential nutritional difficulties, early studies of the dietary and nutritional status of vegetarians (e. g., Hardinger and Stare 1954) and later studies as represented by Taber and Cook (1980) and Bergan and Brown (1980) indicate that vegetarians who maintain a diet with diverse food sources are not nutritionally deficient. Some studies have indicated that the caloric intake particularly among "new" vegetarians was low (Bergan and Brown 1980) and that adult men and women who had recently become vegetarians were generally quite lean. However, measurements of serum vitamins and cholesterol were within the normal range. A well-publicized study by Dwyer and her colleagues of the nutritional status of pre-school vegetarian children indicated that such babies were heavier than the age/sex norms for the general U.S. population but that older children tended to be somewhat leaner and shorter. In general, children on a vegetarian diet are of normal height and weight. The children who were of shorter stature and underweight were following macrobiotic or very restricted vegetarian diets (Dwyer, Andrew, et al. 1980). Nitrogen balance studies of humans on various diets (e. g., Register et al. 1967) indicate that individuals on vegetarian regimens are in appropriate nitrogen balance and are not deficient in amino acids. More specific studies looking at hard-to-get nutrients in the diet of long-term vegetarians have shown that women, both pregnant (King, Stein, and Doyle 1981) and non-pregnant (Anderson, Gibson, and Sabry 1981), show normal levels of iron and zinc in spite of relatively low intake of these minerals (Freeland-Graves, Bodzy, and Eppright 1980; Freeland-Graves, Ebangit, and Bodzy 1980). An additional concern related to the intake of minerals is the use of food items that may be high in phytates, oxalates, and fiber, all of which may interfere with nutrient absorption. I-Iowever, the individuals studied were clearly not suffering from deficiencies in these minerals.


III. Nutritional and Biopsychological Constraints

A study by Marsh et al. (1980) of the cortical bone density of vegetarian and omnivorous women shows no differences in bone mineral mass of 200 lactoovo-vegetarians and 71 omnivores of comparable ages in the third, fourth, and fifth decades of life. These researchers, therefore, propose that calcium and phosphorus absorption and utilization are adequate for these vegetarian women. Dwyer, Miller, and their colleagues (1980) investigated the mental functioning of predominantly vegetarian children. Although the sample was small (N = 28), the results clearly indicate that the children consuming vegetarian diets had an average IQ that was in the 85th percentile of IQ scores of a standardized sample (mean IQ = 116). Studies of the food-related value-orientations, attitudes, and beliefs of vegetarians and non-vegetarians have shown some significant differences (Hufford 1971; Krey 1982; Sims 1978). Vegetarians share: (1) a greater belief in the benefit of health foods; (2) a greater distrust of food processing and additives; (3) a belief that nutrition is important in the maintenance of health; (4) a belief in vitamin supplementation; (5) a strong commitment to weight control; and (6) a belief that they are more knowledgeable about nutrition than non-vegetarians. Tests of knowledge and the use of infonuation sources on nutrition and diet planning did not show significant differences between the American vegetarian and non-vegetarian groups investigated by Sims (1978). Read and Thomas (1983) also found that vegetarians augmented their diet with vitamin and mineral supplements. Eighty-five percent of those studied used food supplements and tended to believe (incorrectly) that naturally occurring vitamins are more nutritious than synthetically produced vitamins. The American Dietetic Association cautiously recommends a non-meatbased or vegetarian diet including eggs and dairy products (1980b). They strongly advocate nutrition education for vegetarians and careful meal planning. In addition, they recommend dietary supplements and the use of fortified products such as soybean milk and flour in a well-balanced multi-item menu. They do not recommend a restricted food intake in tenus of nutrient amounts or specific food items. In summary, individuals on non-restricted vegan or ovo-Iacto-vegetarian diets have adequate nutrient intake, do not show signs of nutritional deficiencies, do not show signs of mental or intellectual impainuent, and may be at reduced risk for cardiovascular diseases and cancers. However, in populations with a more restricted intake, nutritional deficiencies can be common, especially among children and women. Some of these deficiencies are due not to the amount of food ingested but to endogenous chemicals in the foods or to food preparation techniques or to the complicated biological interactions of particular foods. The next section examines some of these biocultural interactions.


9. Biocultural Consequences of Animals Versus Plants

Plant Foods: Dangers and Disorders
Many human toxins occur naturally in plant foods. These toxins are often destroyed by cooking. Exogenous toxins include pesticide and herbicide residues. Some fungi produce toxins (e. g., aflatoxin) that interfere with DNA, RN A, and protein synthesis. This section will briefly review some of the naturally occurring toxins. Of particular interest are the widely distributed oxalates and phytates mentioned previously. The oxalates are found in high concentration in members of the cabbage family, in spinach, beet greens, carrots, peas, rhubarb, mushrooms, and also in tea and cocoa (Committee on Food Protection 1973). These substances bind soluble potassium, sodium, and calcium and make them unavailable for use by the body. Oxalates are also involved in the production of calcium-based kidney stones. Oxalates are heat-sensitive, as are most other toxins, including the phytates. The phytates are found primarily in mature cereal grains, but nuts, legumes, potatoes, com, strawberries, and figs are also sources. Phytates form insoluble complexes with metals-copper, zinc, iron, magnesium, and so on-so that they become unavailable for metabolic functions. Plant polyphenols include phenolic acids, flavanoids, and tannins. They are widely distributed in the leaves, stems, roots, flowers, fruits, and seeds of many plants. These substances, like the goitrogens, can pass into animal products like milk and be ingested by humans as secondary consumers. Phenolic acids and tannins can reduce protein digestibility (Committee on Food Protection 1973). Goitrogens are found primarily in plants of the cabbage family, including turnips, kale, and brussels sprouts. These substances bind iodine. The biocultural consequences of a low iodine-high goitrogen diet are presented below. A number of alkaloids may interfere with the digestion and absorption of nutrients, and some staple foods contain varying amounts of poisons, such as arsenic in potatoes and cyanide in manioc. The potato alone contains some 150 distinct chemical substances, and orange oil has 42 (Coon 1973). The legumes and pulses (e. g., black beans, kidney beans, soybeans, lentils) are some of the best plant protein sources, yet they contain harmful substances called lectins that cause red blood cells to agglutinate and can destroy the walls of intestines, leading to decreased nutrient absorption. Kidney beans also have a substance that inhibits starch digestion. Lectins are destroyed by cooking, and the germination of beans reduces their concentration. Lectins also occur in foods like tomatoes that are frequently consumed without cooking. Finally, foods contain a number of psychoactive agents. For example, substantial amounts of the neurotransmitter serotonin are found in bananas, plantains, and pineapples. The flavor-enhancer monosodium glutamate (MSG),


III. NutritIonal and Biopsychological Constraints

made from beets, is a variant of the neurotransmitter glutamic acid (Lieberman n. d.) Glutamic acid may be responsible for the symptoms in Chinese restaurant syndrome. Perhaps fortunately, many of these neurotransmitters do not cross the blood-brain barrier into the central nervous system. These psychoactive components make up only one of the many classes of drugs derived from plant sources in traditional pharmacopeias. Many techniques have been developed to reduce or eliminate food toxins. Dry storage kills fungi and molds, and the desiccation or drying of foods may eliminate toxins and other potential pathogens. Exposure to light and heat, acid or alkaline environments, fermentation, germination, cooking, and freezing all change the chemical composition of foods (see Table 9.5). Selective breeding for reduced levels of toxins, soil conditions, temperature, humidity, and other environmental factors alter nutrient composition. Certain combinations of food can act synergistically to detoxify harmful chemicals. An example of a significant processing technique involves the preparation of bitter manioc (Katz 1980). Manioc (Manihot esculenta) contains cyanate, which is bound to glycosides distributed throughout the tuber. Any physical damage to the manioc releases a potent glycosidase, which immmediately acts to free small amounts of cyanide (H CN). The processes of squeezing and cooking the manioc release these toxins and volatilize the H CN into the air. These processes are traditionally used by South American Indians. Manioc is also consumed in sub-Saharan Africa, where, in addition to its role in the diet, it has a more complex relationship with the hemoglobin variant HbS, or sickle cell hemoglobin. The cyanate that remains in the manioc is associated with a breaking of the chemical bonds that cause the sickling of red blood cells. This has the advantage of decreasing the risk of anemia due to HbS. At the same time, cells that maintain a sickle shape are more resistant to the malarial parasite. Katz (1980) proposes that the rapid spread of manioc in sub-Saharan Africa over the last 400 years is due to the ability of this crop to reduce the frequency and severity of sickle cell anemic crises. Malaria has been a potent selective agent in human evolution, changing not only the frequency of genes that alter hemoglobin but also the genes for the red cell enzyme glucose-6-phosphate dehydrogenase (G6PD). A deficiency in this enzyme is common in malarial endemic areas, where the deficiency confers some anti-malarial advantage. However, people with G6PD deficiency are susceptible to a potentially lethal attack of a hemoagglutination or of favism caused by the ingestion of the fava bean (Katz 1980; see also Chapter 5). Another important example of the biocultural consequences of the ingestion of potentially pathogenic substances is found in the work of Greene (1977). He investigated the relationship between goiter, goitrogenic plants, and social organization in highland Ecuador. Greene found that a number of indigenous


9. Biocultural Consequences of Animals Versus Plants

plants contained goitrogens, which bind dietary iodine and make it unavailable to the organism. Moreover, at high altitudes people are far from the traditional source of iodine, sea salt. In the two communities studies by Greene, the prevalence of goiter was 52.8 percent and 69.7 percent. In general, these populations had poor supplies of high-quality protein. Many communities contained a significant proportion of cretins, neurologically deficient and behaviorally limited individuals. Cretinism occurs when thyroid function is limited during development. Greene (1977) succinctly documents the severe reduction in the transmission of cultural knowledge when a significant proportion of the population is intellectually and physically impaired. The economic and political ramifications of the subordinate-impaired sub-population's dependence on the dominant-healthy sub-population are starkly portrayed. As with favism and lactase deficiency, an underlying genetic variable is related to the distribution pattern of goiter. This is the genetically inherited ability to taste the manufactured chemical compound phenylthiocarbamide (PTC). PTC is related to the naturally occurring goitrogens. Individuals who are taste-sensitive-that is, those who have inherited a dominant allele for PTC tastingare more likely to avoid or to reduce their intake of bitter-tasting plants containing goitrogens. Non-tasters-individuals who cannot taste this bitter substance-will consume larger quantities of goitrogenic plants. Greene noted a positive correlation between non-tasting and goiter. The final example also shows genotype-diet or genetotrophic interactions. Jt\s in the cases above, food items that are regularly consumed components of dietaries may cause adverse biological reactions-we have noted above, for example, the increased risk of developing cataracts in milk-drinking societies (Simoons 1981). Celiac disease, or gluten sensitivity, is found in populations that are on the margins of wheat-cultivating areas. Populations who are longtime consumers of wheat and other cereal products have a low prevalence of celiac disease. People with celiac disease are sensitive to the substance gliadin, which is a fraction of the gluten in cereal grains. These people suffer from intestinal damage, malabsorption, wasting, and a high frequency of gut and lymphoid cancers (Simoons 1981). Gluten is found in wheat and rye and in lesser concentrations in barley and oats. Triticum wheat has the highest percentage of protein that is gliadin (40 to 60 percent) and therefore causes the lTIOSt severe reactions in celiacs. Therapy involves the consumption of a gluten-free diet. The mechanism by which symptoms are produced by gliadins is not known (Simoons 1981). Although there is no single gene that can account for celiac disease, there is a genetic marker that is part of the immune system. Celiac disease is related to the human leukocyte antigen HLA-B8. This serves as a good genetic marker in U.S. and northwestern European groups. Eightyeight percent of adult celiacs in the United States and England have HLA-B8,


III. Nutritional and Biopsychological Constraints

compared with a frequency of 22 to 30 percent in the general population. HLAB8 appears to be linked to some unidentified gene or gene complex that is selected against in wheat- or cereal-eating populations. The above examples of genetotrophic adaptations indicate the tremendous nutritional and metabolic variability found in contemporary members of our species. These genetotrophic adaptations represent an extreme fonn of adjustment to selective pressures. Humans, like other animals, more readily adapt by changing behavior or modifying the environment rather than changing themselves. The most time-consuming and inflexible fonn of change is genetic. Short- and long-tenn physiological changes offer an intennediate level of flexibility, whereas cultural and behavioral changes can be mustered readily and have the greatest flexibility. As omnivores, our dietary diversity provides both pleasures and pitfalls.

There is great diversity in human nutrient needs and even greater diversity in human diets. The first fact has, in part, accounted for the continuing assessments and revisions of the RDAs. It is also clear that human populations have undergone and continue to undergo physiological modifications in response to "new" foods. These accommodations are often both cultural and biological. Accommodations have been made to nutrients from both animal sources (lactose, high meat diets) and plants sources (wheat, goitrogen-containing plants, soybeans, maize). Yet our variable and flexible metabolisms are being strained by rapid change. New Westernized lifestyles and dietaries are associated with new diseases like New World syndrome, cancers, and cardiovascular disease. Our genetic adaptations are designed for diets containing a wide range of plant and animal foods, moderate energy intakes, a fair amount of dietary fiber, and an energetically more costly lifestyle than is pursued today by most people. As Stini (1971) and many others have indicated, perhaps 99 percent of our evolution has occurred within a hunting, gathering, and scavenging niche. It has allowed for increases in body size and brain development leading to a number of behavioral and physiological characteristics that make our species unique. With the adoption of a more sedentary mode of life approximately ten thousand years ago, a simplification of the diet took place as a preoccupation with crop and animal husbandry reduced the opportunity for hunting and gathering. Usually the diet was and is based on a starchy staple (Robson and Wadsworth 1977). As humans specialized in different environments, the digestive, absorptive, metabolic, and excretory functions of different groups underwent differential selection, leading to the enormous variability in nutritional processes


9. Biocultural Consequences of Animals Versus Plants

and needs at the population and individual level that has been explored in this chapter. The macro- and micro-nutritional excesses and deficiencies of diets high in either animal or plant foods have been presented. Even lean meat, one of our cultural superfoods, can disrupt nonnal metabolic processes when too few calories or not enough fat and carbohydrates are consumed with it. On the other hand, even relatively unbalanced protein sources, if subjected to special treatments (e. g., the lime preparation of maize) and eaten with other complementary plant protein, can provide the same amino acids as lean meat. Humans are ingenious in their approach to solving dietary problems. Of particular importance has been the cooking of foods. Heat, although it may destroy some nutrients, also destroys pathogens and toxins, makes grains palatable, enhances the bioavailability of nutrients, and improves the texture of meat. Our legacy of metabolic and behavioral adaptability has allowed us such abundant dietary choices that whether nutrients come from plants or animals is less important than whether one obtains adequate amounts. The biocultural conditions that create this flexibility are phenomenally diverse.

References Cited
Abrams, H. L. 1980 Vegetarianism: An Anthropological/Nutritional Evaluation. Journal of Applied Nutrition 32: 53-87. Alcantara, E. N., and E. W. Speckman 1976 Diet, Nutrition and Cancer. American Journal of Clinical Nutrition 29: 1035-47. American Dietetic Association 1980a Improving the Amino Acid Profile of Vegetable Proteins. Journal of the American Dietetic Association 76: 141. 1980b Position Paper on the Vegetarian Approach to Eating. Journal of the

American Dietetic Association 77:61-68.
Anderson, B. M.; R. S. Gibson; and]. H. Sabry The Iron and Zinc Status of Long-Term Vegetarian Women. American 1981 Journal of Clinical Nutrition 34: 1042-48. Anderson, ]. W., and K. Ward 1978 Long-Term Effects of High Carbohydrate, High Fiber Diets on Glucose and Lipid Metabolism: A Preliminary Report on Patients with Diabetes. Diabetes Care 1: 77-82. Bergan, ]. G., and P. T. Brown 1980 Nutritional Status of "New" Vegetarians. Journal of the American Dietetic Association 76: 151-55.


III. Nutritional and Biopsychological Constraints

The Energetics of Obesity. Medicine and Science in Sports and Exercise 15:32-40. Committee on Dietary Allowances, Food and Nutrition Board 1980 Recommended Dietary Allowances. 9th rev. ed. Washington, D. C.: National Academy of Sciences. Committee on Food Protection, Food and Nutrition Board 1973 Toxicants Occurring Naturally in Foods. 2d ed. Washington, D. C.: National Academy of Sciences/National Research Council. Committee on Nutrition, American Academy of Pediatrics 1977 Nutritional Aspects of Vegetarianism, Health Foods and Fad Diets. Pediatrics 59: 460-64. Coon, J. 1973 Toxicology of Natural Food Chemicals: A Perspective. In Toxicants Occurring Naturally in Food, 2d ed., pp. 573-91. Washington, D. C.: National Academy of Sciences/National Research Council. Desor, ]. A.; L. S. Greene; and O. Miller Preference for Sweet and Salty in 9-15-Year-Old and Adult Humans. 1975 Science 190:686-87. Draper, H. H. 1977 The Aboriginal Eskimo Diet in Modern Perspective. American Anthropologist 79: 309-16. Dubos, R. 1980 Nutritional Ambiguities. Natural History 89(7): 14, 16, 20-21. Dwyer, ]. T.; E. M. Andrew; ]. Valadian; and R. B. Reed 1980 Size, Obesity and Leanness in Vegetarian Preschool Children. Journal of the American Dietetic Association 77: 434-39. Dwyer, ]. T.; L. G. Miller; N. L. Arduino; E. M. Andrew, W. N. Dietz; ]. C. Reed; and H. B. C. Reed 1980 Mental Age and 1. Q. of Predominantly Vegetarian Children. Journal of the American Dietetic Association 76: 142-47. Flynn, M. 1978 The Cholesterol Controversy. Contemporary Nutrition 3:3-4. Food and Agricultural Organization (FA 0) 1970 Amino Acid Content of Foods and Biological Data on Proteins. FAG Nutrition Studies, no. 24. Rome: F AO. 1973 Energy and Protein Requirements. F AO Nutrition Meeting, Series no. 52. Rome: FAO. Food and Agricultural Organization/World Health Organization (F A 0 / WHO) 1973 Energy and Protein Requirements. Report of the Joint FAG / WHO Ad Hoc Expert Committee. Technical Report Series no. 522. Geneva: WHO. Freeland-Graves, ]. H.; P. W. Bodzy; and M. Eppright 1980 Zinc Status of Vegetarians. Journal of the American Dietetic Association 77:655-61.

Bray, G. A. 1983


9. Biocultural Consequences of Animals Versus Plants

Freeland-Graves, ]. H.; M. L. Ebangit; and P. W. Bodzy 1980 Zinc and Copper Content of Foods Used in Vegetarian Diets. Journal of the American Dietetic Association 77: 648-54. Gleibennann, L. 1973 Blood Pressure and Dietary Salt in Human Populations. Ecology ofFood and Nutrition 2: 143-56. Greene, L. 1977 Hyperendemic Goiter, Cretinism, and Social Organization in Highland Ecuador. In Malnutrition, Behavior and Social Organization, L. Greene, ed., pp. 55-94. New York: Academic Press. Guthrie, H. 1975 Introductory Nutrition. 3d ed. Saint Louis: C. V. Mosby. Hardinger, M. G., and F. Stare 1954 Nutritional Studies of Vegetarians, I: Nutritional, Physical and Laboratory Studies. Journal of C finical Nutrition 2: 73-82. Harris, M. 1979 Our Pound of Flesh. Natural History 88(7):30-36. Harris, R., and E. Kannas 1975 Nutritional Evaluation and Food Processing. 2nd ed. Westport, Conn.: AVI. Harrison. G. G. 1975 Primary Lactase Deficiency: A Problem in Anthropological Genetics. American Anthropologist 77: 128-35. Hufford, D. Organic Food People: Nutrition, Health and World View. Keystone 1971 Folklore Quarterly 16(4): 179-84. Jackson, A. Aminoacids: Essential and Non-Essential? Lancet 1: 1034-37. 1983 Jenkins, D. ]. 1982 Lente Carbohydrate: A Newer Approach to the Dietary Management of Diabetes. Diabetes Care 5:634-41. Katz, S. H. 1982 Food, Behavior, and Biocultural Evolution. In The Psychobiology of Human Food Selection, L. M. Barker, ed., pp. 171-88. Westport, Conn.: AVI. Katz, S. H.; M. L. Hediger; and L. A. Valleroy 1975 The Anthropological and Nutritional Significance of Traditional Maize Processing Techniques in the New World. In: Biosocial Interrelations in Population Adaptation, E. S. Watts, F. E. Johnston, and G. W. Lasker, eds., pp. 195-234. The Hague: Mouton. Katz, S. H., and S. Schall 1979 Fava Bean Consumption and Biocultural Evolution. Medical Anthropology 3:459-76. King, ].; T. Stein; and M. Doyle 1981 Effect of Vegetarianism on the Zinc Status of Pregnant Women. American Journal of C finical Nutrition 34: 1049-55.


III. Nutritional and Biopsychological Constraints

Krey, S. H. 1982 Lappe, F. M. 1975 Lieberman, L. n. d. Liebman, B. 1983

Alternate Dietary Lifestyles. Primary Care 9: 595-603.

Diet for a Small Planet. Westminster, Md.: Ballantine Books. S. Nutrition, Behavior and Adaptation. Manuscript.
Are Vegetarians Healthier Than the Rest of Us? Nutrition Action, June, pp. 8-11.

Liener, I. 1974 Toxic Constituents ofAnimal Food Stuffs. New York: Academic Press. Luft, F.; M. Weinberger; and C. Grim 1982 Sodium Sensitivity and Resistance in Normotensive Humans. American Journal of Medicine 72: 726-36. McCarron, D.; C. Morris; H. Henry; and]. Stanton 1984 Blood Pressure and Nutrient Intake In the United States. Science 224: 1392-98. McCracken, R. D. 1971 Lactase Deficiency: An Example of Dietary Evolution. Current Anthropology 12:495-517. Marsh. A.; T. V. Sanchez; O. Mickelsen; ]. Keiser; and G. Mayor 1980 Cortical Bone Density of Adult Lacto-Ovo-Vegetarian and Omnivorous Women. Journal of the American Dietetic Association 76: 148-51. Moskowitz, H. W.; V. Kumaraiah; K. N. Sharma; H. L. Jacobs; and S. D. Sharma 1975 Cross-Cultural Differences in Simple Taste Preferences. Science 1:1217-18. Multiple Risk Factor Intervention Trial Research Group (MRFITRG) 1982 Multiple Risk Factor Intervention Trial: Risk Factor Changes and Mortality Results. Journal of the American Medical Association 248: 146577. National Center for Health Statistics (N C H S) 1979 Fats, Cholesterol and Sodium Intake in the Diet of Persons 1-74 Years: United States Advance Data 54: 1-12. D HEW Pub. no. 80-1250 Hyattsville, MD.: Department of Health, Education and Welfare. National Livestock and Meat Board Exploring the Unknown: Meat, Diet and Health. Chicago: NLMB. 1983 Neel, James U. 1962 Diabetes Mellitus: A Thrifty Genotype Rendered Detrimental by "Progress"? American Journal ofHuman Genetics 14:353-62. Nutrition Search, Inc. 1975 Nutrition Almanac. New York: McGraw-Hill. Read, M., and D. C. Thomas 1983 Nutrient and Food Supplement Practices of Lacto-Ovo-Vegetarians. Journal of the American Dietetic Association 82:401-4. Reed, P. B. 1980 Nutrition: An Applied Science. St. Paul: West.


9. Biocultural Consequences of Animals Versus Plants

Register, U. D., et al. 1967 Nitrogen Balance Studies in Human Subjects on Various Diets. American Journal of Clinical Nutrition 20: 753-59. Rennie, M., and R. Harrison 1984 Effects of Injury, Disease and Malnutrition on Protein Metabolism in Man. Lancet 3:323-25. Robson, ]. R. K., and G. R. Wadsworth 1977 The Health and Nutritional Status of Primitive Populations. Ecology of Food and Nutrition 6: 187-202. Rosch, P. 1983 Stress, Cholesterol and Coronary Heart Disease. Lancet 2:851-52. Rosenbaum, R. 1976 Crunch. Esquire, July, pp. 57-60, 114. Rozin, E., and P. Rozin 1981 Some Surprisingly Unique Characteristics of Human Food Preferences. In Food in Perspective, A. Fenton and T. Owen, eds., pp. 243-52. Edinburgh: John Donald. Schaefer, O. H. 1971 When the Eskimo Comes to Town. Nutrition Today 6:8-16. 1981 Eskimo (Inuit). In Western Diseases: Their Emergence and Prevention, H. Trowell and D. Burkitt, eds., pp. 113-28. Cambridge: Harvard University Press. Scrimshaw, N. S., and V. R. Young 1976 The Requirements of Human Nutrition. Scientific American, September, pp. 51-64. 1978 Biological Variability and Nutrient Needs. In Progress in Human Nutrition, vol. 2, S. Margen and R. Ogar, eds., pp. 102-31. Westport: AVI. Simoons, F. ]. 1969 Primary Adult Lactose Intolerance and the Milking Habit: A Problem in Biological and Cultural Interrelations, I: Review of the Medical Research. American Journal ofDigestive Diseases 14:819-36. 1970 Primary Adult Lactose Intolerance and the Milking Habit: A Problem in Biological and Cultural Interrelations, II: A Cultural Historical Hypothesis. American Journal ofDigestive Diseases 15: 695-710. 1982 Geography and Genetics as Factors in the Psychobiology of Human Food Selection. In The Psychobiology of Human Food Selection, L. M. Barker, ed., pp. 205-24. Westport: AVI. 1981 Celiac Disease as a Geographic Problem. In Food, Nutrition and Evolution, D. Walcher and N. Kretchmer, eds., pp. 179-99. New York: Masson. Sims, L. 1978 Food-Related Value-Orientations, Attitudes, and Beliefs of Vegetarians and Nonvegetarians. Ecology ofFood and Nutrition 7:23-35. Sirton, C., and M. Lovati 1984 Soybean Protein Diet: Experimental and Clinical Studies on the Mode of Action and Therapeutic Efficacy. In Diet, Diabetes and Atherosclerosis,


III. Nutritional and Biopsychological Constraints

G. Possa, P. Micossi, A. Catapano, R. Paoletti, eds., pp. 241-53. New York: Raven Press. Snowdon, D., and R. Phillips 1985 Does a Vegetarian Diet Reduce the Occurrence of Diabetes? American Journal ofPublic Health 75: 507-12. Snowdon, D.; R. Phillips; and G. Fraser 1984 Meat Consumption and Fatal Ischemic Heart Disease. Preventive Medicine 13: 490-500. Speth, ]., and K. Spielmann 1983 Energy Source, Protein Metabolism, and Hunter-Gatherer Subsistence Strategies. Journal ofAnthropological Archaeology 2: 1-31.

Stini, W.
Evolutionary Implications of Changing Nutritional Patterns in Human Populations. American Anthropologist 73: 1019-30. Taber, L. A., and R. A. Cook 1980 Dietary and Anthropometric Assessments of Adult, Omnivores, FishEaters and Lacto-Ovo-Vegetarians. Journal of the American Dietetic Association 76:21-29. Trowell, H., and B. D. Burkitt Western Diseases: Their Emergence and Prevention. Cambridge: Har1981 vard University Press. Tufts University Newsletter 1984a Cholesterol: Is the Verdict Finally In? Special Report. Tufts University Diet and Nutrition Newsletter 2: 3-50. 1984b The Estrogen-Calcium Gap. Tufts University Diet and Nutrition Newsletter 2: 1. Waterlow, ]. C., and P. R. Payne 1975 The Protein Gap. Nature 258: 113-17. Weiss, K.; R. Ferrell; and C. Hanis 1984 A New World Syndrome of Metabolic Diseases with a Genetic and Evolutionary Basis. Yearbook ofPhysical Anthropology 27: 153-78. West, R., and O. B. Hayes 1968 Diet and Serum Cholesterol Levels: A Comparison Between Vegetarians and Nonvegetarians in a Seventh-Day Adventist Group. American Journal of C linical Nutrition 21: 853-62. Wilmsen, E. 1978 Seasonal Effects of Dietary Intake on Kalahari San. San Federation Proceedings 37:65-72. Wing, E. S., and A. B. Brown 1979 Paleonutrition: Method and Theory in Prehistoric Foodways. New York: Academic Press. 1971


Pre-State Foodways: Past and Present
GRANTED THAT THERE ARE INNATE BIOPSYCHOLOGICAL REstraints on human diets, we are still left with the problem of accounting for the enormous variability offoodways and their divergence and convergence during the course ofcultural evolution. The essays in Part IV deal with these diachronic and synchronic variations in foodways among pre-state societies, from the perspective of archaeology (Cohen and Yesner), from the perspective of ethnohistory (D. Harris), and from the perspective of ethnography (Winterhalder, Hawkes, Johnson and Baksh, and Good). The archaeological issue addressed by Cohen is the transition from hunting-gathering to agricultural modes ofproduction with the attendant "revolutionary" consequences for human foodways. Yesner addresses an equally important but less well known aspect of the evolution of human foodways-namely, the increased utilization of maritime resources at the end of the Pleistocene that paralleled and sometimes preceded the shift to domesticated plant and animal species. Both Cohen and Yesner invoke aspects of optimal foraging theory in relation to resource depletion and population growth in order to explain why apparently abundant species and simple food production technologies were neglected for tens of thousands ofyears prior to the Neolithic. In addition, Cohen presents data that indicate that the transition to neolithic modes of food production was actually accompanied by a deterioration in nutritional standards-a paradoxical consequence, but one that is not incompatible with optimal foraging theory. Two chapters devoted entirely to optimal foraging theory follow. Winterhalder defends the basic epistemological and theoretical principles that underlie this approach, reviews some of its applications to the explanation of diet breadth and foraging patterns among several contemporary foraging groups, and offers new models that incorporate opportunity costs and sensitivity to risks. Hawkes then addresses the question of why Kalahari !Kung men do not spend as much time as the men of the Ache ofParaguay in hunting, and why Ache women do not spend as much time as !Kung women in gathering. OJ special interest in these two


IV. Pre-State Foodways

papers is the central theoretical significance Hawkes attaches to foodways selected for by their consequence for reproductive success and Winterhalder's explicit denial that the analysis of optimality in foraging behavior is inextricably bound to reproductive success models. In a different vein, David Harris describes the broad-spectrum subsistence regimen of northern Australian Aborigines. This is of special interest not only for its ecological interpretation of ethnohistorical sources, but for its treatment of cannibalism as an integral feature of a foraging economy. The final two essays in Part IV are concerned with the foodways of two Amazonian tropical forest village societies, the Machiguenga ofPeru and the Yanomami of Venezuela. Johnson and Baksh find that despite the existence of some aversions that apparently cannot be reduced to ecologically utilitarian costs and benefits, the Machiguenga exploit the possibilities of their technology and habitat in an overwhelmingly pragmatic fashion. Although Good's study of Yanomami hunting patterns does not address the adaptive efficiency of their foodways, it does provide valuable new information on the extent to which preoccupation with scarce animal food results in semi-migratory settlement patterns and intra-village factionalism associated with village fissioning.



The Significance of Long-Term Changes in Human Diet and Food Economy
common set of nutritive needs as well as a set of limited physiological capabilities for intake, absorption, storage, and excretion and a sensory apparatus that guides food choice. All appear in part anachronistically geared to the ancestral primate environment, with some modifications from the period of early hominid adaptation. We also inherit from our ancestors the capacity for flexible, widely ranging, omnivorous solutions to nutritive problems and a habit, already observable in other primates, of forming culturally determined food lists - that is, of learning food habits, which are determined as much by the experience of our parents and peers as by our physiological limits. Given this heritage, the range of human dietary habits can be viewed as an aggregation of localized, idiosyncratic cultural solutions to common nutritive problems (minor, recently evolved differences in food tolerance and utilization, such as variations in lactase or sucrase production or the tasting/non-tasting of goitrogenic compounds notwithstanding). Each "successful" solution must approximate adequate and balanced nutrition for enough members of the population, particularly those in critical roles or at critical life periods, to live and reproduce; and it is the requirements for this success that selectively limit the range of idiosyncratic variation in diet. 1 But it is clear that to be "successful" a solution need not provide equally for all and need not provide what we would consider a "healthy" diet. Moreover, it is clear that any group's choices serve a number of functions other than good nutrition and thus that nutritional values are always compromised by competing functions in the design of food strategies. Indeed, part of a group's successful adaptation may lie in the nutritional deprivation of some of its members (Cassidy 1980, 1983, 1984). Local variations in strategy occur in part because of local variations in the distribution and timing of potential resources. "Optimal foraging" theorists have, during the past few years, given us a handle with which to approach


IV. Pre-State Foodways

these local variations by asserting a series of propositions about common animal behavioral tendencies through which the adjustment of common human needs and predilections to different environmental circumstances can be assessed (Winterhalder and Smith 1981). But optimal foraging theory can, at best, recognize limited parallels among varying local economies. It can identify the expression of basic needs and of common adaptive strategies and common behavioral principles, such as the "principle" of maximizing calorie returns for labor. In this way it can explain the "systematic" fraction of any food economy; however, it can at best isolate, but not explain, the non-systematic (or "noise") fraction of any local adaptation resulting from competing cultural purposes or the baggage of taste, tradition, and habit. The noise fraction represents, I suspect, a far larger portion of human behavior than most anthropologists are willing to allow. Much behavior witnessed much of the time, whether through archaeological or ethnographic techniques, is, I believe, noise at the level of most of our explanatory paradigms. We cannot hope to explain the specifics of what people eat through any means other than a study of the histories of individual cultures. What we can do is recognize some limited parallels across systems that are expressible as statistical trends and tendencies across populations. Optimal foraging theory is in fact a set of such statements recognizing common solutions to certain adaptive problems across species and across living human cultures. At least two other major sets of predictive statements have been offered describing alternative sets of limited parallels among human food economies (sets that complement one another and the theorems of optimal foraging rather than being contradicted by either). These are statements about regional or latitudinal variations (such as the latitudinal predictions of Lee 1968) and statements of temporal or sequential variation (Cohen 1977; Hayden 1981a, 1981b). The latter, which provide the focus of this paper, suggest that amidst the variety of human cultures, and cutting across regularities of both latitude and local optimization, there are recognizable trends in human food preferences over time.

Human Food Economies Through Time
Recent syntheses of world prehistory (see Cohen 1977; Fagan 1983; Hayden 1981a, 1981b) largely agree about what the archaeological record appears to show concerning changes in human food habits through time. The earliest fully human groups in most world regions appear to have been selective in their use of biomes, preferring open savanna-like biomes, regardless of latitude, to more heavily vegetated ones; and these early populations appear to have focused


10. The Significance of Long-Term Changes in Human Diet and Food Economy

relatively heavily (though in unknown proportions) on the taking of large game and a limited range of vegetable foods. Smaller game, non-mammalian (particularly avian) fauna, fish, and shellfish, and a broader range of vegetable foods, particularly small seeds and starchy tubers, seem increasingly to have been exploited only in relatively recent prehistory. The roots of these trends are traceable into the Pleistocene in the Old World; the trends culminate in the post-Pleistocene "broad-spectrum revolution." The attention newly paid to small seeds and starchy tubers then seems to have resulted repeatedly in the domestication of crops, whereby such foods were caused to grow in high densities in the vicinity of human settlements, altered genetically to grow and be harvested at human convenience, and further altered to produce enlarged and often detoxified and otherwise unprotected edible parts. Simultaneously, however, human habits were domesticated to the convenience of the reproductive cycle of the plants in question. Human groups were forced (or permitted) to be sedentary, and work habits were altered to fit the growth cycles of the plants. This sequence appears to have been replicated with fair regularity on each of the world's major continents. (Australian prehistory has often been cited as an exception to this rule because it has been said to display an essentially static adaptation, but in at least one recent synthesis [Lourandos 1983], the Australian sequence is said to show a pattern of intensification of resource use very similar to that of other continents.) Significantly, the sequence seems to be replicated whether it is played out over hundreds of thousands of years, as in most of the Old World, or telescoped into thousands, or at most tens of thousands, as in the New. The meaning of this apparent pattern, however, is in dispute on at least two levels. First, it is uncertain to what degree the archaeological record is to be taken at face value or must be modified by considerations of differential preservation and/or biases in the reporting and interpretation of data. Second, even if the evidence has been correctly read, it is uncertain how the sequence is to be interpreted-that is, what it implies in terms of changing human behavior. A number of authors have criticized the sequence of changing food habits described above, suggesting that it is primarily a sequence of archaeological preservation rather than one of actual dietary change (e. g., Dahlberg 1981; Kamminga 1981). They note that most modem human hunter-gatherers (outside extreme latitudes) eat a wide range of (predominantly vegetable) foods; that the range of foods at least occasionally preserved at very early human sites suggests a similar broad spectrum of exploitation; that the items usually considered late additions to the human diet are archaeologically fragile and may simply have been removed from the early part of the record by differential destruction; and that on seacoasts, where much of the recent broad-spectrum revolution is observed, early sites have now largely been destroyed by rising sea levels.


IV. Pre-State Foodways

It is clear that a sufficient record of broad-spectrum food use from early human history does indeed exist to indicate that there has been no simple sequence of pure economic types from the legendary "big-game hunters" of the Middle Pleistocene to the broad-spectrum gatherers at the end of the Pleistocene and after. The evidence of primate omnivory itself (Harding and Teleki 1981) is sufficient to cast the image in doubt, and I would agree that there is clear evidence of fairly omnivorous foraging early in the human record. Yet I would argue that a good case can still be made for a real quantitative trend in human diets over time-as a statistical tendency across diverse economies. I would suggest that our diets have shifted in the ways the record suggests even though that trend may be exaggerated by differential preservation. As I have argued elsewhere (Cohen 1977), the trend occurs not only along coasts but also inland, where sea level, despite its indirect effects on rivers, does not exert the profound and uniform influence on the preservation of archaeological sites that it does on the coast. The trend also occurs in regions of the coast where sea level change can be ruled out (see Chapter 11); it occurs over the relatively short span of New World prehistory, through which differential preservation necessarily is far less important as a selective factor than it is over the long span of Old World prehistory; and it occurs also in patterns of changing tool frequencies (particularly with regard to the frequencies of grinding tools) not subject to differential preservation. Since I initially wrote and presented the arguments above, moreover, a number of new lines of argument have become available that tend to help establish the reality of the trends described. One possible line of evidence, discussed at the Wenner-Gren Foundation Conference on Food Preferences and Aversions, concerns the human ability to absorb dietary iron. Human beings are inefficient absorbers of iron from vegetable sources. Iron deficiency is a common problem among modem human beings and among archaeological populations beginning with the adoption of agriculture (Cohen and Armelagos 1984; see discussion below). This pattern may bespeak an evolutionary heritage somewhat richer in iron sources (and particularly animal products) than are recent agricultural diets. (In much the same manner, our inability to synthesize ascorbic acid is commonly attributed to an earlier stage of our ancestry in an environment rich in vitamin C.) Another, firmer line of argument is provided by optimal foraging theory and its application to living populations. Several such models have suggested that a widening of the foraging base (the broad-spectrum revolution) is a predictable response by organisms to an increase in their own density, the decline of prey species, an improvement in food-getting technology, the deterioration of the environment, or all of these (Christenson 1980; Earle 1980; Hespenheide 1980; Winterhalder 1981). Moreover, specific studies of exploitative efficiency have suggested that in the expansion of the food base of human groups, large


10. The Significance of Long-Term Changes in Human Diet and Food Economy

game animals should be (and commonly are) a preferred resource, whereas small seeds, small game, avifauna, shellfish, and vegetable resources requiring more than minimal processing should be (and commonly are) low-priority resources that foragers take increasingly as returns from hunting decline (Hames and Vickers 1982; Hawkes, Hill, and O'Connell 1982; O'Connell and Hawkes 1981; Perlman 1983). Such observations increase the likelihood that the observed archaeological sequence represents a real pattern of adaptive changes rather than a mere artifact of preservation. A further point is that human skeletal remains themselves display a history of changes throughout the Pleistocene that appear to represent either micro-evolution or plastic responses to changing diets and food economies. Wolpoff (1980) has speculated, for example, that changes in skeletal size and in tooth structure in the Homo line in the Early and Middle Pleistocene reflect readaptation of a foraging ancestral line to a more carnivorous diet at the same time that my own reading of the archaeological data suggests that efficient hunting first emerged (see Cohen 1977). A firmer case can be made for a very widespread reduction in skeletal size and robusticity (Angel 1984; Frayer 1981; Kennedy 1984; Meiklejohn et al. 1984; Smith et al. 1984) that began, at least throughout the Old World, at the end of the Pleistocene. The trend has been alternately considered a response or adaptation to declining nutrition or a result of increased dependence on shellfish or a relaxation of selective pressures on robusticity. But whatever interpretation is selected, very broadly parallel trends in food use and/or food acquisition are probably implicated. A further test of the reality of the apparent economic change is now emerging in the form of trace element and isotope analysis of bones, which has successfully demonstrated short-term changes as well as intra-site variations in diet but which so far has had only limited success in reconstructing dietary changes of any time depth. One published study (Schoeniger 1982) appears to demonstrate a long-term decline (from the Middle Paleolithic to the Mesolithic) in the quantity of meat in human diets in the Middle East. However, this argument has been criticized by others (Smith et al. 1984). If the observed trends in the archaeological record are thus at least partly real (if partly exaggerated by preservation), their significance is still in dispute. In recent years, as indeed through much of the history of the discipline, there have been two major schools of thought concerning these trends. One of these schools celebrates the improvement in human life brought about by the advancement of technology. Childe (1951) is probably the most famous proponent of this position among prehistorians; Hayden (1981a, 1981b) is its most recent major spokesman. Its models tend to eqaute changes in food economy with advances in technology that permitted new resources to be exploited. Others have proposed that the increasingly broad spectrum of resources utilized was more a function of need than of progress. These latter models


IV. Pre-State Foodways

propose that new resources were adopted not so much as technology permitted but as the relative scarcity of preferred resources demanded. The scarcity of preferred resources is in turn seen as a function of the increasing size and density of (or demand generated by) human populations and the decline in the populations of preferred prey species resulting either from the inroads of human predation or from the detrimental effects of post-Pleistocene patterns of climate change (see Binford 1983; Cohen 1977; Harris 1977, 1979). The same issue exists between "stress" and "progress" models of the adoption of agriculture itself. The adoption of agriculture, once considered clearly a technological advance (e. g., Childe 1951), has in recent years itself been viewed by some as an adaptation to stress generated by population growth or a decline in the productivity of wild resources (Cohen 1977; Harris 1977, 1979). The two schools of thought largely agree that increases in the size and density of human populations are associated with changing technology. Where they differ, primarily, is in interpreting the effects of increasing density and new technology on the quality of human life and in characterizing the motives behind economic change. Models emphasizing technological progress (e.g., Hayden 1981a, 1981b) tend to assume that greater numbers of people were supported with increasing ease and with increasing reliability (if not quality) of food supplies and to assume that such technological improvements were positively sought by our forebears. A stress model (Cohen 1977) implies, on the contrary, that economic changes necessitated by population growth or resource decline most commonly resulted in diminished returns for labor and in declining quality and reliability of food supplies. A model offered by Harris (1977, 1979) suggests something of an intermediate pattern in which economic stress motivates change but occasionally promotes changes of sufficient magnitude (such as the adoption of cultivation) temporarily to relieve stress, producing shortlived periods of relative prosperity before population growth initiates a new cycle of stress. One possible means of differentiating between these stress and progress models is to consider human food preferences. Both Harris and I have argued that the meat (and, as we must now concede, even more the fat) of large animals is a preferred food for most human groups, suggesting that the expansion of the diet away from this resource is likely to represent stress rather than progress in the eyes of those making the change. This assertion is supported by two lines of evidence that emerged at the Wenner-Gren conference. First it was argued (see Chapter 4) that although non-human primates commonly consume less animal food than do people, they do nonetheless treat it as a preferred food when available. In addition, there was a consensus among ethnographers in the group that, among modern human beings, the consumption of animal products was highly correlated with social class (and the ability to choose resources), suggesting that human beings tend to consume higher


10. The Significance of Long-Term Changes in Human Diet and Food Economy

amounts of animal food when they can, higher amounts of vegetable food when they must. Both sets of observations suggest that the broadening of the food spectrum away from animal products represents necessity, not choice. The optimal foraging theories described above seem to me to distinguish between progress- and stress-related theories at least for the foraging (preagricultural) part of the sequence. The broadening of the resource base itself is ambiguous in this regard, since a wider resource based is predicted to result either from an increased population of predators (or a decline in prey) or from increased technological capabilities (Winterhalder 1981). However, the studies of resource ranking cited above suggest that, quite independent of taste, preference, and the quality of the nutrient package, the meat of large animals provides a more efficient caloric return for labor than do most other wild resources, even among modem foragers utilizing an essentially mesolithic (postPleistocene) technology. If these results can be extrapolated back into the past (when, if anything, large game was more plentiful and hence more easily taken), they suggest that the broad-spectrum revolution, despite its new technology, was probably accompanied by declining returns for labor (decreasing caloric efficiency). The results thus suggest that necessity, not progress, motivated economic change. (These data, of course, do not directly compare the efficiency of foraging strategies to the efficiency of cultivation, two broadly different economies whose relative labor costs are also widely debated. The results do suggest, however, that cultivation focusing on small seeds and tubers was initially directed at one of the least efficiently exploited and least valued parts of the foragers' repertoire.)

The Evidence of Skeletal Pathology
The last decade has seen explosive growth in the field of skeletal analysis and paleopathology, growth that promises to help resolve many of the issues presented above and already provides substantial evidence on several of them (see Buikstra and Cook 1980; Cohen and Arrnelagos 1984; Huss-Ashmore et al. 1982). The growth in these fields includes not only an enormous increase in the number of studies done, but also a new, ecological, population-oriented (rather than specimen- or pathology-oriented) approach and the appearance of new analytic techniques. Supplementing the traditional identification of specific pathologies, modem skeletal analysis now recognizes and quantifies a number of more general conditions of biological stress. Both quantification itself and the identification of non-specific indicators of stress facilitate comparisons between populations with regard to overall health and nutrition. It is possible to describe populations with regard to the incidence and severity of infection, inflammation,


IV. Pre-State Foodways

and malnutrition, and to combine this description with quantitative analysis of the frequency of growth-disrupting stresses experienced by individuals and analysis of the age distribution of deaths. Using these techniques, it is then possible to begin to evaluate the quality of the diet provided by different food economies; the degree to which different economies (or concomitant features of settlement) exposed their participants to infection, illness, physical stress, and trauma; the frequency of periodic starvation or other growth-disrupting events; and the life expectancy offered to participants. In recognition of this potential, Cohen and Armelagos (1984) invited scholars who possessed such comparative data for hunter-gatherers and farmers in various regions of the world to make comparative assessment, where possible, of populations at four broadly defined economic stages: early (paleolithic or PaleoIndian) hunter-gatherers; late, broad-spectrum (mesolithic or Archaic) huntergatherers; transitional or incipient agricultural populations; and established agriculturalists. We obtained 18 regional sequences of varying degrees of completeness and detail, distributed as follows. There were eight regional sequences from the United States, including one each from California (Dickel et al.), Ohio (Perzigian et al.), Kentucky (Cassidy), Georgia (Larsen), the lower Mississippi (Rose et al.), the Southwest (Palkovich), and two from Illinois (Goodman et al. ; Buikstra and Cook). The New World was further represented by one study from Central America (Panama and Costa Rica by Norr et al.); one from Ecuador (Ubelaker), and two from Peru-Chile (Benfer, Allison). Two sequences were reported from Europe (Western Europe by Meiklejohn; the Mediterranean by Angel), two from the Middle East (the Levant by Smith et al.; IranIraq by Rathbun), one from South Asia (Kennedy), and one from North Africa (Nubia by Martin et al.). (All are listed in the References Cited section with 1984 citations.) The comparison of these sequences poses a number of problems. Each sequence involves comparison of several distinct populations~ thus representing, in effect, a series of snapshots of a continuous process of economic evolution in the region. Moreover, because of the nature of archaeological preservation and recovery, each sequence is composed of a slightly different series of snapshots, so that it is difficult to isolate precisely equivalent stages of transition for comparison. In addition, each sequence involves its own unique series of confounding factors, such as changing climate, the movement of people, or the action of political systems; and in each sequence the interaction of dietary change, population growth, population nucleation, and sedentism (factors that might exert independent effects on health) is unique. The studies also suffer from the fact that skeletal populations are often skewed representations of living groups, and from the fact that different analyses have been performed (or the same analyses have been performed in a different manner) in different regions and often


10. The Significance of Long-Term Changes in Human Diet and Food Economy

on different populations from the same region, making both inter-regional comparisons and inter-population comparison within a region more difficult. Perhaps the most important limitation of the studies as a whole, from the point of view of the questions raised here, is that skeletal remains of truly early, mobile, hunter-gatherers (paleolithic or Paleo-Indian populations) are either unrepresented or are represented only by scattered, often fragmentary finds. Most statements about hunter-gatherers offered here, therefore, reflect relatively late pre-agricultural groups, groups that may have adopted the sedentism, class-stratification, and relatively large population aggregates that are usually considered more typical of farmers. Only tantalizing glimpses are offered of the earlier groups, and conclusions about hunter-gatherer lifestyles must be tempered accordingly, although the existing data often show clear trends that might be extrapolated back into the past. A further problem is that skeletal populations occasionally lack association with sufficient organic archaeological refuse to permit nice discriminations between stages of economic change. In a few instances, associated economies are postulated almost entirely from associated artifacts and site distributions, supplemented by data from skeletons themselves (including data on the frequency of tooth caries or the trace-element and isotope composition of bone that are considered indicative of diet). Even under the best circumstances, quantified estimates of diet are generally unavailable. Given these caveats, the most reliable conclusions drawn from the data are those which relate to the simplest, crudest distinctions. In a simple comparison of hunting and gathering populations with the agricultural populations that came after them, for example, a number of fairly clear trends emerge. But attempts to describe trends in health across subtler transitions-that is, attempts to make finer distinctions among economic stages-rapidly encounter problems of poor sampling or poor resolution.

The Health of Hunter-Gatherers as Compared with Farmers

The clearest major trend in the comparison of hunter-gatherers and farmers concerns the incidence of infection as measured by the frequency of generalized infectious lesions of bone and specific infectious diseases. Twelve of the 18 studies report on the incidence of infection (Allison; Angel; Buikstra and Cook; Cassidy; Goodman et al.; Larsen; Meiklejohn et al.; Norr; Perzigian; Rathbun; Rose et al.; Ubelaker). Almost all conclude that infection was a more


IV. Pre-State Foodways

serious problem for farmers than for their hunting and gathering forebears, and most suggest that this resulted from increasing sedentism, larger population aggregates, and/or the well-established synergism between infection and malnutrition. Conversely, a decline in the incidence of infection among farmers as opposed to hunter-gatherers is reported only by Norr and Rose et al. (and by the latter for only one of three sub-areas in their study.) Rathbun, while reporting a higher rate of infections for neolithic and chalcolithic populations than for hunter-gatherers in Iran, reports, surprisingly, that later Bronze and Iron age populations showed the lowest infection rates. In addition to reporting on generalized lesions of infection, two studies (Buikstra and Cook; Perzigian) note an increase in the frequency of mycobacterial (tubercular) infections identifiable in the skeletons of farming or later populations; and one (Allison), working with mummies, has been able to document an increase in rates of gastrointestinal infections with sedentism and agriculture, although he finds no trend in respiratory diseases.

The studies also suggest fairly consistently that farming resulted in a decline in the overall quality of nutrition. The clearest indicator of this is provided by the incidence of porotic hyperostosis and cribra orbitalia (porosity of the skull and orbits), considered indicative of anemia. Such anemia is most commonly attributed to iron deficiency and may reflect low-meat, high-cereal diets or the competition of parasites such as hookworm for available iron. (In the Old World such lesions may also reflect genetic anemias not directly related to diet but tied to malaria, which in turn has been shown to be responsive to changing human economic strategies.) Sixteen studies (Allison; Angel; Buikstra and Cook; Cassidy; Goodman et al.; Kennedy; Larsen; Martin et al.; Meiklejohn et al.; Norr; Palkovich; Perzigian et al.; Rathbun; Rose et al.; Smith et al.; Ubelaker) estimate rates of porotic hyperostosis, and most (12 of 16) conclude that the lesions appear or increase with farming, suggesting that anemia, whether nutritional, parasite-related, or genetic, is primarily a farmers' disease. The remaining four studies (Allison; Larsen; Martin et al.; Meiklejohn et al.) either provide no hunter-gatherer data or find no trend in their data. Other indicators of chronic malnutrition are more sporadically reported and, in some cases, are less certain of interpretation; but these other indicators seem generally to support a similar conclusion. Four of the studies (Buikstra and Cook; Cassidy; Goodman et al.; and Smith et al.) employ either the cortical thickness or relative length-for-age of long bones to assess nutrition, and all four found that chronic malnutrition was more common among farmers than among their hunting and gathering forebears. Angel, using measurements of


10. The Significance of Long-Term Changes in Human Diet and Food Economy

skull-base height and pelvic-inlet depth as indices of nutrition, found a decline in the overall quality of nutrition from the Paleolithic through the Neolithic in the Mediterranean region. Several of the studies also suggest that the smaller stature and reduced skeletal robusticity among farmers as compared with hunter-gatherers might be indicative of declining quality of nutrition (Angel for the Mediterranean; Kennedy for India; Larsen for Georgia; Meiklejohn for Europe), but not all of the studies agree on the significance of this indicator, and not all show similar trends. For example, Meiklejohn et a1. report a decrease in stature from Paleolithic through Mesolithic to Neolithic in Europe, followed by a rebound, but cautiously concludes that the trend might reflect either dietary stress or altered activities. Smith et a1. show a decline in stature from paleolithic to mesolithic groups in Israel but argue against considering this an index of declining nutrition, since there are no accompanying stress indicators. They note a rebound in stature in the Neolithic. Rose et a1. suggest that an increase in size in the Caddoan region of the lower Mississippi valley might reflect an increase in available protein with farming while noting that a decline in size, reflecting declining nutrition, occurred in the lower Mississippi valley proper.

Episodic Stress
One of the most interesting and, in some ways, most problematic sets of data is provided by indicators of episodic stress sufficient to produce disruptions in growth. Two such indicators were commonly reported, with surprisingly conflicting results. Harris lines (growth disruption lines in long bones) are reported in seven studies comparing hunter-gatherers and farmers (Allison; Buikstra and Cook; Cassidy; Goodman et a1.; Perzigian et a1.; Rathbun; Rose et a1.). Most conclude that the lines are more common in hunter-gatherers than in the farmers who followed them, suggesting that farming reduced the frequency of episodic stress. Rathbun, in contrast, reports an increase in these lines through time; Allison finds regional differences but no trend among different economies; and Cook notes that the lowest frequency of Harris lines occurred in her transitional group. Linear enamel hypoplasias and enamel micro-defects of the teeth tell a very different story, however. Ten studies (Allison; Angel; Buikstra and Cook; Cassidy; Goodman et a1.; Kennedy; Perzigian et a1.; Smith et a1.; Ubelaker; and, very tentatively, Rathbun) report that the frequency or the severity of this indicator of growth disruption increased in farming and later populations in comparison with hunter-gatherers, suggesting more frequent and more severe episodes of stress in later groups. One study finds no trend (Rose et aI., for one of their three sub-areas), and the other studies do not report this index. In other


IV. Pre-State Foodways

words, no study finds a decline in the frequency of this index over time. Three studies (Cassidy; Perzigian et al.; and Smith et al.) also report that hypoplasias of deciduous teeth (indicating episodes of pre-natal stress presumably reflecting poor maternal health) were more common in later agricultural groups than in hunting and gathering groups. The apparently contradictory pattern of Harris lines and enamel defects may be resolved in various ways. One possibility is that the two represent stresses of differing etiology. Both are considered general indicators of stress sufficient to cause growth disruption, but in neither case is the range of possible stressing agents well defined. Hence, it is possible that the contradictory trends indicate a trade-off of one form of stress for another. Cassidy suggests that minor, regular periods of hunger among hunter-gatherers may have been traded for the more irregular and more severe stresses of farming life, involving bouts of infection and the more serious famine associated with crop failure. It was also pointed out in several of the papers and in discussion that enamel hypoplasias are almost certainly the more reliable indicator of the two. Harris lines are subject to subsequent erasure during growth, whereas enamel hypoplasias are not. Moreover, Harris lines record not so much the cessation of growth as the subsequent compensatory acceleration of growth. It can be argued, therefore, that the lower frequency of Harris lines in farming populations reflects chronic malnutrition and the resulting failure of the acceleration phase as much as the actual incidence of growth arrest. (This interpretation is supported by a recent study of laboratory monkeys [Murchison et al. 1983] that demonstrates that chronic malnutrition does indeed prevent the body from developing Harris lines in response to stress episodes.) At best, therefore, the two indicators taken together suggest that no relaxation of episodic stress resulted from the adoption of farming in most regions. They may indicate the trade-off of one type of stress for another; and they may indicate a net increase in such episodes that is not recorded in Harris lines because of a concomitant decline in overall "background" levels of nutrition.

Physical Stress
The data on physical stress (work stress and physical injury) are more mixed but seem to argue, on balance, for a reduction in physical stress and therefore a probable reduction in workload associated with the adoption of agriculture. Lower levels of arthritis and skeletal robusticity are reported for farmers in comparison with hunter-gatherers in several studies (Kennedy; Larsen; Meiklejohn; Perzigian; Smith et al.). Others report more complex patterns (Rathbun; Rose et al.) or note an increase in rates of arthritis with the adoption


10. The Significance of Long-Term Changes in Human Diet and Food Economy

of farming (Goodman et al.). Cassidy suggests, however, that the arthritis she observed was of infectious etiology. It should be pointed out that both arthritis and muscular robusticity are likely to reflect the severity of peak or intermittent demand on muscles and joints rather than simply the number of hours of work involved in the two economies. The data thus suggest that, generally, such peak demand was greater for hunter-gatherers than for farmers, but they do not necessarily indicate that farming resulted in a reduction in the time invested in the food quest. The incidence of trauma, both accidental and violence-related, shows a similarly mixed picture, generally decreasing with the adoption of agriculture in some study regions (Meiklejohn et al.; Perzigian et al.; Rathbun) and increasing in others (Goodman et al.; Kennedy).

Life Expectancy
Perhaps the most interesting results concern the changes in life expectancy associated with the transition to agriculture. Thirteen of the regional studies provide data (of varying quality) bearing on this issue. Of these, 10 either conclude, or provide fragmentary data suggesting, that the life expectancies of farming populations were lower than those of preceding hunter-gatherers. Working with relatively good, well-controlled samples (N = 114, 224, 219, respectively) from a single well-defined location, Goodman et al. suggest that there was a progressive decline in life expectancy for all age classes as huntergatherers first adopted and then intensified agriculture in Illinois. Buikstra and Cook, from large and well-controlled samples (N = 800+, 500+, 336), describe a decline in life expectancy with the adoption of agriculture, followed by a recovery of life expectancy for adults (but not for children) late in the sequence as agriculture was intensified. Cassidy finds that a maize-farming group in Kentucky (N = 296) had a lower life expectancy for both sexes and for all age groups than did an earlier hunting and gathering group (N = 295). Ubelaker notes that life expectancy at birth was relatively high in an early huntergatherer group (N = 192); rose slightly in what might be an agricultural group (N = 199), which, however, still exhibited a low caries rate, leading him to question their degree of reliance on agriculture; and then fell sharply in an intensive agricultural group (N = 30) before a partial rebound occurred (N = 435). He suggests that there was a trend toward increasing infant mortality through his sequence. Life expectancies at more advanced ages (E 5 and E 15 ) fluctuated more irregularly over time, but the earliest clearly hunting and gathering population had life expectancies at both ages comparable to the average of later groups.


IV. Pre-State Foodways

Larsen, while cautioning that available samples (N = 152, 177) are unlikely to be truly representative because both showed too high a percentage of survivors to adulthood, nonetheless notes that the age distribution of the available sample of farmers was lower than that of the earlier hunter-gatherers in Georgia. He reports a higher percentage of individuals dead in all age classes below 25 for farmers and a higher percentage of hunter-gatherers in all age classes over 25, with the exception of the 30- to 35-year age bracket. Even if sub-adult mortality is discounted, the pre-agricultural adults had the higher average age at death. In the conference version of their paper, Rose et al., working with samples of similar size, suggested that in the Caddoan region of the lower Mississippi, hunter-gatherers had a lower probability of dying either as children or as young adults than did later farmers. In the lower Mississippi valley proper, they also suggest, the lowest probability of dying was associated with a hunter-gatherer group. Working with more fragmentary or scattered data, a number of other studies report similar trends. Angel notes a decrease in adult life expectancy from Paleolithic (N = 53) to Mesolithic (N = 120) and Neolithic (N = 106) for males in the Mediterranean. For females he reports an increase in adult life expectancy from Paleolithic (N = 53) to Mesolithic (N = 63), followed by a decline to levels at or below that of the Paleolithic for the Neolithic (N = 200). Kennedy notes that the aggregate pre-agricultural sample from India (N = ± 100) display higher ages at death than do later agricultural populations. Rathbun, who claims no trend because his samples are so small, cites figures on average age at death of known adult specimens that are higher for pre-agricultural individuals (N = 9) than for early agricultural (neolithic) groups (N = 69). According to his tables, 10 small samples of neolithic burials all had average ages at death for adults below those of the Shanidar Neanderthals and the pre-neolithic Hotu population (although the Hotu sample may not be representative). And Allison suggests that childhood mortality was lowest in his earliest population (although recovery bias is probably involved in this sample). These latter studies, which suffer from very poor sampling of early hunting and gathering populations, mean relatively little individually, but collectively and in combination with the larger and better-controlled samples discussed above, they begin to suggest a fairly widespread pattern of declining longevity with the adoption of farming. Certainly they at least challenge the prevailing reverse presumption. Three studies, however, suggest an increase in life expectancy with farming. Smith et al. report a general increase in life expectancy from late pre-agricultural (Natufian) populations in the Levant through the Bronze age; Perzigian et al. cite a comparison suggesting better survivorship for an agricultural (N = 44) group in Ohio than for pre-agricultural groups (N = 1,327), but they noted that the later, smaller sample almost certainly underestimated sub-adult mortality, accounting at least in part for the difference in life expec-


10. The Significance of Long-Term Changes in Human Diet and Food Economy

tancy. Figures provided by Norr might also be read as suggesting greater survivorship to adulthood in agricultural populations compared with pre-agricultural populations in Panama; but she adds several disclaimers about both the size (N == 87, 28, 32) of her samples and probable age biases related to known burial practices. Taken as a whole, these indicators suggest an overall decline in the qualityand probably in the length-of human life among farmers as compared with earlier hunter-gatherer groups. Farmers seem fairly commonly to have suffered more infection and more chronic malnutrition than their forbears living on wild resources, and they seem to have suffered as many or more episodes of growth-disrupting stresses and to have had reduced life expectancies. On the other hand, there is evidence that farming often placed fewer physical demands on the skeleton (in terms of peaks of mechanical stress) than did hunting and gathering. It should be reiterated here that the hunter-gatherer/farmer comparison masks a number of separate adaptive shifts whose importance in accounting for differences in health is not always clear. The effects of new foods per se, of farming as an activity, of sedentism, of the dispersal and nucleation of settlement, and of altered politics should all be considered separately, but these separate effects are necessarily lumped together in any simple comparison (see especially Rose et al. 1984).

Biological Stress and the Broad-Spectrum Revolution
The data available for assessing conflicting interpretations of the broad-spectrum revolution are fewer (because of the relative dearth of early huntergatherers), and the results are necessarily far more tentative. The clearest pattern in the studies, already alluded to, is a decline in stature from paleolithic to mesolithic populations, which occurred through much of the Old World (as reported by Angel; Kennedy; Meiklejohn et al.; and Smith et al.). Unfortunately this widespread trend toward reduced body size at the end of the Pleistocene can be read two ways: either as a sign of a general decline in nutrition or as a function of reduced workloads and reduced muscular, physical stress. A concomitant decline in arthritis and robusticity (Kennedy; Meiklejohn et al.; Smith et al.) seems to argue for the latter explanation. A concomitant increase in other indicators of malnutrition would support the former one; and such is reported in one region (by Angel for the Mediterranean). Such evidence is not noted in the other regions (where the indices of malnutrition used by Angel were not employed), but if reduced body size resulted from selection for survival with a poor diet rather than from plastic individual growth response, sur-


IV. Pre-State Foodways

viving individuals might not show signs of distress in their own skeletons even though the population was evolving to adapt to nutritional deprivation. Indicators of growth disruption (Harris lines and enamel hypoplasias) should hold the key to distinguishing between "stress" and "progress" models of the broad-spectrum revolution, but the data are extremely scanty. Comparative Harris line data generally are not reported among Old World populations from this time period, and only Smith et al. report on the relative frequencies of enamel hypoplasias (noting no trend). An earlier study (Brothwell 1963), with an extremely small sample from Europe, notes an increase in the frequency of hypoplasias from the Paleolithic to the Mesolithic. In the New World there are again no data comparing Paleo-Indians with archaic and later groups. Dickel et al. (1984), however, report on indicators of health among California Indians, who, although they never became farmers, did display a pattern of changing economic exploitation of wild resources similar to the evolution of broad-spectrum exploitation elsewhere. They report that Harris lines declined through the sequence, possibly suggesting improved homeostasis as a broader spectrum of resources was utilized, but they note steady or increasing rates of enamel hypoplasia and declining life expectancies for most age classes through time. In a study of pre-agricultural populations on the coast of Peru, Benfer (1984) finds improvement in a number of indices of health (Harris lines, stature, life expectancy) in a population that moved first toward increased use of vegetable resources and then back toward increasing use of marine fauna. Benfer also notes that the group appeared to be degrading its resources. In sum, there are not yet sufficient data (nor are the existing data sufficiently consistent) to permit interpretation of the meaning of the broad-spectrum revolution in terms of human health. There is no clear pattern either of increasing stress or of improving homeostasis.

The Harris Hypothesis: Temporary Amelioration of Stress Through Adoption of Agriculture
The Harris (1977; 1979, and Chapter 2) hypothesis that the adoption of agriculture would have resulted in temporary improvement in the quality of life meets with only mixed success. In the Old World, Smith et al. report two trends that appear to support this hypothesis. Early neolithic populations in the Levant display an increase in stature over preceding mesolithic (Natufian) populations, and analyses of strontium content in the bones of these same populations suggest that there was an increase in the consumption of animal products by the neolithic group, presumably as a result of animal domestication. At the


10. The Significance of Long-Term Changes in Human Diet and Food Economy

same time, however, these authors also suggest that chronic malnutrition was first evidenced more generally in the Neolithic in the form of thinned cortical layers of long bones, and they note that rates of enamel hypoplasia are higher for neolithic groups than for earlier ones. In Rathbun's data from Iran, moreover, the Neolithic appears to be a period of peak stress in terms of rates of infection, arthritis, and at least one indicator of malnutrition, porotic hyperostosis. In Rathbun's figures, later populations seem to have adjusted more successfully to farming than did the Neolithic populations. Angel's figures for the Mediterranean suggest a similar pattern. Several indicators of malnutrition in this study suggest that nutrition declined into and through the Neolithic and that a rebound in nutrition occurred only beginning with the early Bronze age. In Europe, Meiklejohn et al. note that the decline in stature that began in the Mesolithic extended into the Neolithic, with rebound beginning only late in the Neolithic. Martin et al. suggest that early farming populations were healthier than later farmers, but they offer no pre-farming groups for comparison. In the New World, Cassidy notes that in Kentucky a (poorly reported) transitional (incipient agricultural) group may have enjoyed good health in contrast to both the hunter-gatherers before them and (more strikingly) the farmers who followed, as the Harris hypothesis would suggest. Perzigian also finds some suggestion of an improvement in health in Ohio with the initial adoption of farming. He notes that there was an increase in stature with the transition to farming followed by a later decrease; rates of enamel hypoplasia did not increase significantly until the later agricultural periods. Rose et aI., discussing various portions of the Mississippi valley, find a complex picture that varies from region to region. For the Caddoan region they note a temporary decline in rates of arthritis associated with the adoption of farming, suggesting that the new economy temporarily reduced the workload. But they also note that both rates of infection and rates of porotic hyperostosis peaked in this transitional group. In their sequences from Illinois, Buikstra and Cook suggest that the early farming population showed the generally poorest health and nutrition; and Goodman et al. find a steady decline in most health indicators without any indication of a temporary improvement. In Latin America, Norr notes a steady increase in rates of porotic hyperostosis but a reduced rate of infection in the early agricultural group. And Ubelaker finds that a group generally regarded as early farmers displayed various indications of relatively good health (compared with people before them, and especially people after). But the rate of caries in this group led him to question whether they were truly agricultural. Harris's expectation of a temporary amelioration in health and nutrition following the adoption of farming is therefore only occasionally or partially fulfilled (in part perhaps because of the difficulty of identifying and differentiating the various stages of economic change with precision and consistency). Equally common is a pattern in which early farmers display poor health in comparison


IV. Pre-State Foodways

with populations both before and after them. It may be that the amelioration Harris predicts is offset by the stress associated with economic change per se and that adoption of a new economy results in poor health until appropriate adjustments are made. It is interesting that there is no clear difference between Old and New World populations with regard to the stresses experienced by early farmers. The presence of large domesticable animals in the Old World might lead us to expect that the adoption of farming would be less nutritionally stressful there than in the New World, where no large animals were domesticated. But the data, at present, allow no such simple distinction.

Beginning in the Middle Pleistocene, the archaeological records of the various regions of the world display broadly similar patterns. Groups appear first to focus relatively heavily on the exploitation of large mammals and then to expand the range of resources utilized to include increasing amounts of small game, avian and other non-mammalian fauna, fish and shellfish, and small seeds and other vegetable material. These trends, which culminate after the end of the Pelistocene, are now commonly referred to as the "broad-spectrum revolution." The attention paid to small seeds and vegetable sources of calories then appears to have resulted in various regions of the world in the domestication of crops and in the commitment of human populations to dependency upon them as primary sources of calories. Although our perception of these trends may be affected by differential preservation in the archaeological record, which may exaggerate the differences in economy between different periods, various lines of evidence suggest that these patterns do not result solely from differential preservation. In particular, there is evidence that the archaeological sequence corresponds to known patterns of resource preference among primates and modern human beings; and it appears that the sequence conforms fairly well to "optimal foraging" predictions (and related ethnographic observations) about the sequence in which various types of resources should be and are utilized by optimizing populations. These same data seem to suggest that both the broad-spectrum revolution and the subsequent focus on small seeds and other vegetable sources of calories by farmers were more a function of necessity (resulting from either increasing human population or declining resources) than a function of improved technological capacity. The meaning of these trends in terms of the quality of human life, nutrition, and life expectancy can now be assessed by means of skeletal pathology. The


10. The Significance of Long-Term Changes in Human Diet and Food Economy

data from skeletal analysis suggest fairly consistently that the adoption of farming commonly resulted in declining health and nutrition, steady or increasing rates of episodic stress, and declining life expectancies in comparison with earlier hunting and gathering, although it often resulted, too, in reduced physical stress. The data do not yet allow us to determine what the broad-spectrum revolution implied for human health. The extant data simply show no clear pattern either of increasing stress or of improved nutritional homeostasis. Similarly, among archaeological farming populations there is no clear, consistent trend in health and nutrition when incipient and established farmers are compared.

1. In contrast with the stricter cultural materialism of the editors of this volume, I hold a view of human cultural evolution adopted from Campbell (1965). Campbell asserted that cultural evolution parallels biological evolution in being a two-stage process. New variations are created or introduced to a population by a series of mechanisms (invention and diffusion in cultural evolution, mutation and gene flow in the biological context). These variations are then secondarily channeled by differential survival or selection. The second, selective phase qf this process eliminates dysfunctional or competitively unsuccessful variations and thus imposes limits on the potential range of variation. Surviving cultures (or parts of cultures), like surviving organisms, must therefore display some minimum degree of adaptive design, the degree being dependent on the stiffness of the competition. But the largely random nature of the original creative forces and the existence of competing sets of selective forces suggest that, within the limits imposed by selection, a good deal of random, non-adaptive, or even mildly dysfunctional variation is possible and is to be expected. Cultural evolution theory can no more predict all of the features of cultural design than the rules of biological evolution can predict the design of specific organisms.

References Cited
Allison, M. 1984 Paleopathology in Peruvian and Chilean Populations. In Paleopathology at the Origins of Agriculture, M. N. Cohen and G. Armelagos, eds., pp. 515-29. New York: Academic Press. Health as a Crucial Factor in the Changes from Hunting to Developed Farming in the Eastern Mediterranean. In Paleopathology at the Origins

Angel, ]. L. 1984


IV. Pre-State Foodways

ofAgriculture, M. N. Cohen and G. Armelagos, eds., pp. 51-73. New York: Academic Press.
Benfer, R. 1984 The Challenges and Rewards of Sedentism: The Preceramic Village of Paloma, Peru. In Paleopathology at the Origins of Agriculture, M. N. Cohen and G. Armelagos, eds., pp. 531-58. New York: Academic Press.

Binford, L. R. 1983 In Pursuit of the Past. New York: Thames and Hudson. Brothwell, D. 1963 The Macroscopic Dental Pathology of Some Earlier Populations. In Dental Anthropology, D. R. Brothwell, ed., pp. 271-88. Oxford: Pergamon. Buikstra, ]. 1984 The Lower Illinois River Region: A Prehistoric Context for the Study of Ancient Diet and Health. In Paleopathology at the Origins ofAgriculture, M. N. Cohen and G. Armelagos, eds., pp. 215-69. New York: Academic Press. Buikstra, ]., and D. C. Cook 1980 Paleopathology: An American Account. Annual Review ofAnthropology 9:433-70. Campbell, D. 1965 Variation and Selective Retention in Sociological Evolution. In Social Change in Developing Areas, H. R. Barringer et aI., eds., pp. 19-49. Cambridge, Mass.: Schenkman. Cassidy, C. M. 1980 Benign Neglect and Toddler Malnutrition. In Social and Biological Predictors ofNutritional Status, Physical Growth and Neurological Development, L. S. Greene and F. Johnston, eds., pp. 109-39. New York: Academic Press. Commentary on "Food: Past, Present and Future" by N. S. Scrimshaw. 1983 In How Humans Adapt, D. Ortner, ed., pp. 253-57. Washington: Smithsonian Institution. 1984 Skeletal Evidence for Prehistoric Subsistence Adaptation in the Central Ohio River Valley. In Paleobiology at the Origins of Agriculture, M. N. Cohen and G. Armelagos, eds., pp. 307-45. New York: Academic Press. Childe, V. G. 1951 Man Makes Himself New York: Mentor. Christenson, A. L. 1980 Changes in the Human Niche in Response to Population Growth. In Modeling Change in Prehistoric Subsistence Economies, T. K. Earle and A. L. Christenson, eds., pp. 31-72. New York: Academic Press. Cohen, M. N. 1977 The Food Crisis in Prehistory. New Haven: Yale University Press. Cohen, M. N., and G. Armelagos (eds.) 1984 Paleopathology at the Origins of Agriculture. New York: Academic Press.


10. The Significance of Long-Term Changes in Human Diet and Food Economy

Dahlberg, F. (ed.)

1981 Woman the Gatherer. New Haven: Yale University Press. Dickel, D., et al. 1984 Central California: Prehistoric Subsistence Changes and Health. In Paleopathology at the Origins of Agriculture, M. N. Cohen and G. Armelagos, eds., pp. 439-61. New York: Academic Press. Earle, T. 1980 A Model of Subsistence Change. In Modeling Change in Prehistoric Subsistence Economies, T. K. Earle and A. L. Christenson (eds.), pp. 1-29. New York: Academic Press. Fagan, B. People of the Earth. 4th ed. Boston: Little, Brown. 1983 Frayer, D. W. Body Size, Weapon Use and Natural Selection in the European Upper 1981 Paleolithic and Mesolithic. American Anthropologist 83: 57-73. Goodman, A., et al. 1984 Health Changes at Dickson Mounds, Illinois (A. D. 950-1300). In Paleopathology at the Origins of Agriculture, M. N. Cohen and G. Armelagos, eds., pp. 271-305. New York: Academic Press. Hames, R. B., and W. T. Vickers 1982 Optimal Diet Breadth Theory as a Model to Explain Variability in Amazonian Hunting. American Ethnologist 9: 358-78. Harding, R. S. 0., and G. Teleki 1981 Omnivorous Primates: Gathering and Hunting in Human Evolution.
New York: Columbia University Press. Harris, M. R. 1977 Cannibals and Kings. New York: Vintage. 1979 Cultural Materialism. New York: Random House. Hawkes, K.; K. Hill; and]. F. O'Connell 1982 Why Hunters Gather: Optimal Foraging and the Ache of Eastern Paraguay. American Ethnologist 9: 379-98. Hayden, B. 1981a Research and Development in the Stone Age: Technological Transitions Among Hunter-Gatherers. Current Anthropology 22: 519-48. 1981b Subsistence and Ecological Adaptations of Modern Hunter/Gatherers. In Omnivorous Primates: Gathering and Hunting in Human Evaluation, R. S. O. Harding and G. Teleki, eds., pp. 344-422. New York: Columbia University Press. Hespenheide, H. A. 1980 Ecological Models of Resource Selection. In Modeling Change in Prehistoric Subsistence Economies, T. H. Earle and A. L. Christenson, eds., pp. 73-78. New York: Academic Press. Huss-Ashmore, R., et al. 1982 Nutritional Inference from Paleopathology. Advances in Archaeological Method and Theory 5:395-474. Kamminga, ]. 1981 Comment on B. Hayden. Current Anthropology 22: 535-36.


IV. Pre-State Foodways

Kennedy, K. A. R. 1984 Growth, Nutrition and Pathology in Changing Paleodemographic Setting in South Asia. In Paleopathology at the Origins of Agriculture, M. N. Cohen and G. Armelagos, eds., pp. 169-92. New York: Academic Press. Larsen, C. 1984 Health and Disease in Prehistoric Georgia: The Transition to Agriculture. In Paleopathology at the Origins ofAgriculture, M. N. Cohen and G. Armelagos, eds., pp. 367-92. New York: Academic Press. Lee, R. D. 1968 What Hunters Do for a Living; or, How to Make Out on Scarce Resources. In Man the Hunter, R. B. Lee and I. DeVore, eds., pp. 30-48. Chicago: Aldine. Lourandos, H. 1983 Intensification in Australian Prehistory. Paper read to International Congress of Anthropological and Ethnological Sciences, Vancouver. Martin, D., et a1. 1984 The Effects of Socioeconomic Change in Pehistoric Africa: Sudanese Nubia as a Case Study. In Paleopathology at the Origins of Agriculture, M. N. Cohen and G. Armelagos, eds., pp. 193-214. New York: Academic Press. Meiklejohn, C., et a1. 1984 Socioeconomic Change and Patterns of Pathology and Variation in the Mesolithic and Neolithic of Western Europe: Some Suggestions. In Paleopathology at the Origins of Agriculture, M. N. Cohen and G. Armelagos, eds., pp. 75-100. New York: Academic Press. Murchison, M. A., et a1. 1983 Transverse Line Formation in Protein Deprived Rhesus Monkeys. Paper presented to the annual meeting of the Paleopathology Association, Indianapolis. Norr, L., et a1. 1984 Prehistoric Subsistence and Health Status of Coastal Peoples from the Panamanian Isthumus of Lower Central America. In Paleopathology at the Origins of Agriculture, M. N. Cohen and G. Armelagos, eds., pp. 463-90. New York: Academic Press. O'Connell, ]. F., and K. Hawkes 1981 Alyawara Plant Use and Optimal Foraging Theory. In Hunter-Gatherer Foraging Strategies: Ethnographic and Archeological Analysis, B. Winterhalder and E. A. Smith, eds., pp. 99-125. Chicago: University of Chicago Press. Palkovich, A. Agriculture, Marginal Environments and Nutritional Stress in the Pre1984 historic Southwest. In Paleopathology at the Origins of Agriculture, M. N. Cohen and G. Armelagos, eds., pp. 425-38. New York: Academic Press. Perlman, S. Optimum Diet Models and Return Rate Curves: A Test on Martha's 1983


10. The Significance of Long-Term Changes in Human Diet and Food Economy

Vineyard Island. In Ecological Models in Economic Prehistory, G. Bronitsky, ed., pp. 115-67. Anthropological Research Papers, no. 29. Tempe: Arizona State University. Perzigian, A. ].


Prehistoric Health in the Ohio River Valley. In Paleopathology at the Origins ofAgriculture, M. N. Cohen and G. Armelagos, eds., pp. 347-66. New York: Academic Press.

Rathbun, T.


Paleopathology and Economic Change in Iran and Iraq. In Paleopathology at the Origins ofAgriculture, M. N. Cohen and G. Armelagos, eds., pp. 137-67. New York: Academic Press. Paleopathology and the Origins of Maize Agriculture in the Lower Mississippi Valley and Caddoan Culture Areas. In Paleopathology at the Origins ofAgriculture, M. N. Cohen and G. Armelagos, eds., pp. 393-424. New York: Academic Press. Diet and the Evolution of Modem Human Form in the Middle East.

Rose, ]., et a1.


Schoeniger, M.


American Journal ofPhysical Anthropology 58:37-52.
Smith, P., et a1. 1984 Archaeology and Skeletal Evidence for Dietary Change During the Late Pleistocene/Early Holocene in the Levant. In Paleopathology at the Origins ofAgriculture, M. N. Cohen and G. Armelagos, eds., pp. 101-36. New York: Academic Press. Ubelaker, D. 1984 Prehistoric Human Biology of Ecuador: Possible 'remporal Trends and Cultural Correlations. In Paleopathology at the Origins ofAgriculture, M. N. Cohen and G. Amlelagos, eds., pp. 491-513. New York: Academic Press. Winterhalder, B. 1981 Optimal Foraging Strategies and Hunter-Gatherer Research in Anthropology: Theories and Models. In Hunter-Gatherer Foraging Strategies: Ethnographic and Archeological Analysis, B. Winterhalder and E. A. Smith, eds., pp. 13-35. Chicago: University of Chicago Press. Winterhalder, B., and E. A. Smith (eds.)

Wolpoff, M.

Hunter-Gatherer Foraging Strategies: Ethnographic and Archeological Analyses. Chicago: Aldine. Paleoanthropology. New York: Knopf.




Life in the "Garden of Eden": Causes and Consequences of the Adoption of Marine Diets by Human Societies

IN EXAMINING THE BIOCULTURAL BASIS OF HUMAN FOOD PREFERences and aversions, it is important to document not only short-term idiosyncratic aspects of human foodways, but also longer-term historical trends underlying the acceptance and rejection of food types. Both the causes and the consequences of such major transformations in human foodways need to be assessed. It is in this spirit that this paper attempts to address the question of the causes and consequences of the relatively late adoption of seafood diets by human societies, using some current research being conducted in Maine and Alaska. The historical fact that maritime resources were not exploited until relatively late in the prehistoric record has been recently discussed by Osborn (1977) and others, and has attracted a general consensus of agreement. Very few coastal sites predate the late Upper Paleolithic in the Old World, even in areas (such as England) with emergent shorelines in which earlier sites are not likely to have been erased by subsequent coastal erosion. Those sites that do predate the late Upper Paleolithic show only very scanty evidence of the use of coastal resources (e. g., at Terra Amata; see de Lumley 1969) or certainly much less use of those resources than in Upper Paleolithic or later times at the same location (e.g., at Klasies River mouth in South Africa; see Singer and Wymer 1982; Volman 1978). The overall pattern of these sites suggests that a real commitment to maritime lifeways did not precede late Upper Paleolithis times. Similarly, in the New World, early occupants (Paleo-Indians and hypothetical pre-Clovis occupants) have not been found to have used marine resources, and in most parts of North America a real commitment to maritime lifeways postdates the mid-Holocene (Yesner 1983). What is it in the nature of marine re-


IV. Pre-State Foodways

sources, then, that prevented their exploitation until relatively late in prehistoric times, and what conseqeunces did this use have on the development of human societies? Binford (1983) suggests that coastal resources have been characterized by many archaeologists, including myself, as a "Garden of Eden" in which there is "plentiful food just for the picking," providing "a setting for the beginnings of sedentary life." Binford finds this proposition distasteful because it leads to what he terms a "historical problem": "If aquatic resources are to be considered Gardens of Eden, why did early populations apparently not realize this fact?" This, in turn, seems to lead to the implication that, in terms of the seacoast, "some peoples were 'more perceptive' or 'smarter' ... while others ignored its self-evident advantages" (Binford 1983: 202). If, indeed, such mentalistic notions are to be rejected in favor of ones based on natural selection, then we are left asking two questions. First, is the Garden of Eden model accurate, or must we replace it with a model that suggests that coastal resources are "second rate" (Osborn 1977) in order to explain why coastal environments were not initially exploited in either the Old or the New World? If neither model should prove to be accurate, however, what alternative forces may have led to a failure to exploit marine resources until relatively late? A second paradox that needs to be addressed has to do with the set of constraints imposed upon a society by its resource base. Binford (1983) notes that some have used the Garden of Eden principle to suggest why huntergatherers have rejected agriculture in some places (e. g., coastal California), whereas in other places (e. g., coastal Peru), the sedentism associated with intensive harvesting of seafoods has been viewed as a causative agent underlying the development of agriculture. In addition, the range of political organization that has been attributed to maritime hunter-gatherers ranges from bandlevel societies (e. g., among the Yahgan of Tierra del Fuego) to "big-man" types of societies (e. g., among coastal Californians or New Englanders), hereditary chiefdoms (e. g., among Aleuts or Northwest Coast Indians), and even incipient state-level societies (e.g., in coastal Peru). How, then, do the constraints of the productive process help us to understand these seemingly divergent consequences of maritime adaptation for the development of cultural complexity?

The "Garden of Eden" Versus the "Second-Rate Resource" Model
Recently a number of archaeologists-including myself (Yesner 1981)-have argued that optimal foraging theory presents a robust model that allows one to explain why certain resources are or are not included in the diet (as well as how


11. Life in the "Garden of Eden"

much time or energy is devoted to the exploitation of those resources). Following the tenets of this theory, an item is included in the "optimal diet" if it is "valuable" in energetic terms, usually expressed as the difference between its per-unit caloric yield and its caloric cost (including search, pursuit, and "handling" costs such as retrieval and processing). Whether or not an item is to be included in the diet has little to do with its own abundance but depends solely on its ranking as a resource in terms of "value" and the abundance of more "valuable" or highly ranked resources. As a result, a lower-ranked resource will often not be exploited until more highly ranked resources decline in abundance. Thus, Hayden (1981) has noted that human beings rarely eat field mice in spite of their abundance, because the per unit cost of exploiting such resources is extremely high. Nevertheless, hunter-gatherers living in marginal environments such as deserts do end up eating a lot of rodents, lizards, snakes, insects, and grubs, not necessarily because they are more available, but because more "valuable" resources (such as large mammals) may be less available. Hill (1982) and O'Connell and Hawkes (1981) have recently shown that hunting in general produces a per unit benefit that far outweighs that of plant collecting, and Simms (1983) has also shown that large-mammal hunting produces an energy benefit that outweighs small-mammal hunting (which, in tum, generally outweighs plant collection). If, as Hill (1982) has suggested, optimal foraging theory can be used to show why human beings originally became hunters-particularly big-game hunters-then it could potentially be used to suggest why marine resources were not used until late: that is, because they are high-cost, low-benefit, "secondary" resources. The major problem with this approach, of course, involves estimating the costs and benefits associated with marine resource exploitation. Furthermore, lumping of seafoods into a single category obscures the analysis of costs and benefits. In terms of cost, for example, it is clear that complex technology of any sort is rarely required to collect shellfish or other marine invertebrates (which are successfully harvested with rakes, forks, hooks, or spades), but even here, species that burrow or live in high-energy environments may be more "costly." The same applies to fish, which range from more easily exploited nearshore varieties to "costlier" (although often high-yielding) deep-sea varieties. Sea birds may be planktonic feeders with generally smaller ranges, which are more easily caught, or diving fish-eaters, which are more difficult to exploit. Sea mammals are considered to be "high-cost" resources that generally require complex technology to obtain, but this is certainly much more true of migratory whales than basking seals. Cross-cutting all of these differences is the fact that nearly all marine resources are aggregated, or obtainable in large concentrations under certain conditions, which allows reduced search and retrieval costs. Shellfish beds, fish schools, anadromous fish runs, sea bird colonies, ice leads, and sea mammal


IV. Pre-State Foodways

rookeries all offer conditions in which these resources can be taken at reduced cost. Archaeological evidence from numerous coastal regions, furthermore, suggests that within each category of sea life, less costly harvesting methods tend to be replaced over time by more costly ones. For example, in coastal California, early emphasis on picking surface shellfish, hooking nearshore schooling fish, and clubbing seals at haulouts was replaced by digging clams, trolling for offshore solitary fish species, and (in some areas) actively hunting large sea mammals (Meighan 1983; Tartaglia 1976). In the Old World, longterm sequences of coastal occupation-such as at Klasies River mouth-reveal a gradual shift from strandlooping to more "costly" marine hunting techniques (Singer and Wymer 1982) and certainly Arctic prehistory can be interpreted as a developmental sequence in which there was increased hunting of more "costly" species such as large baleen whales. On the Maine coast, where I have been working most recently, one shift in late prehistoric times apparently involved increased exploitation of a small coastal mammal called the "sea-mink," which is now extinct (Loomis 1911). Clearly the "cost" of exploiting marine resources depends upon the class of resources and often upon the individual species involved. However, the use of boats for "central-place" foraging among marine hunters reduced many search and particularly retrieval costs for a wide variety of resources, and many improvements in maritime technology were largely directed toward increasing the range of boats and thus reducing mobility costs per unit of time. Much of the real cost of marine hunting, however, involves not so much the technology as the social relations of production, ranging from the cost of maintaining family or clan territories in order to restrict access to areas where resources aggregate (anadromous fishing locales, productive shellfish beds, sea mammal haulouts) to the cost of maintaining the structure of whaling crews and the ceremonial apparatus necessary for effective whale hunts. If the energetic "costs" of marine resource exploitation are difficult to calculate, the "benefits" are equally so. Here again it is difficult to deal with marine resources in any kind of monolithic fashion. For example, some seafoods, such as shellfish, are notoriously low in calories (about 0.6 to 1.0 kilocalories per gram), but sea mammals are certainly equal to terrestrial mammals in the caloric content of their flesh, and the fact that they often come in very large packages increases the benefit of exploiting them per unit of effort. Marine fish vary greatly on both counts, ranging from about 0.95 to 3.38 kilocalories per gram (Arimoto 1962). The caloric differences observed among seafoods are related in part to their fat content, since fats have twice the caloric density of either proteins or carbohydrates. Sea mammals are obviously high in fat content, ranging from about 22 to 24 percent fat (Heller and Scott 1962). Shellfish, at the other end of the spectrum, range from only 0.4 to 1.3 percent fat, while marine fish range wide-


11. Life in the "Garden of Eden"

ly from less than 1 percent fat among species like haddock to 25 percent among species like tuna (Arimoto 1962; Heller and Scott 1962). Speth and Spielman (1983) have recently emphasized the dietary role of fats in energy production among hunter-gatherers, particularly in places where cold stress is prevalent all or part of the year, and at times when carbohydrate energy sources are unavailable. In addition, triglycerides allow for transport of fat-soluble vitamins, while essential fatty acids serve as precursors for the manufacture of phospholipids and prostaglandins with honnone-like functions, as well as cholesterol, which is required for steroid and bile salt fonnation (Shodell 1983). An important distinction, however, is that marine foods generally contain larger amounts of unsaturated fatty acids such as eicosapentaenoic acid (E PA), which not only produce as much energy as saturated fatty acids but also tend to have a depressive effect on blood cholesterol levels (Toyama and Kaneda 1962). Thus, it has recently been suggested that the 13 prostaglandins produced by EPA have protected coastal peoples who consume substantial amounts of marine fats (including fish oils) from high blood pressure, atherothrombotic disorders, and other cardiovascular disease (Fischer and Weber 1984). However, problems with blood clotting and other physiological processes may occur if dietary sources of cholesterol are insufficient. Furthermore, marine fatty acids tend to have an antagonistic effect on tocopherol (vitamin E), and an important fatty acid, linoleic acid, is generally less prevalent in marine diets (Toyama and Kaneda 1962). Clearly latitudinal variation is important in this whole picture. Although marine species diversity-particularly fish species diversity-is generally higher in tropical regions, many of these species have a lower fat and caloric content. This is particularly true of reef fish, although other fish and sea mammals may also be affected. In higher latitudes, although marine species diversity is lower, fattier sea mammal and fish species are more available, including both anadromous and catadromous fish. (Salmon, for example-an anadromous fishaverages 17 percent fat content, and eels-catadromous fish-average 18 percent fat content; these species produce about 2.25 kilocalories per gram of body weight [Arimoto 1962].) In the Arctic, shellfish tend to disappear above the southern boundary of pack ice, since ice scouring greatly affects their ability to reproduce and maintain their habitat. These latitudinal effects are compounded by the fact that "plants edible by man decrease in diversity and abundance poleward from the tropics through the temperate zones" (Harris 1982: 19). This has led to an inverse latitudinal trend in human utilization of the two most abundant types of high-calorie foods-terrestrial plants and marine animals-such that occupants of different latitudinal zones tend to substitute one resource for the other. If the use of particular seafoods in different latitudinal zones is viewed in historical perspective, additional patterns emerge. For example, in northern


IV. Pre-State Foodways

temperate and Subarctic zones such as the Maine-Maritimes region or the Pacific coast from southwestern Alaska to Washington state, sea mammals were exploited earlier than anadromous fish or shellfish. One might well argue that the caloric benefits for these populations outweighed the considerable costs associated with sea mammal hunting and that shellfish were not exploited until pressures on more "valuable" sea-mammal resources necessitated the inclusion of shellfish in an optimal diet. Optimal foraging theory thus appears helpful in explaining such phenomena, but it is important to be careful that what we are observing is not simply the result of environmental change-for example, the fact that shellfish beds could not initially form because of rapid postglacial sea-level rise, particularly in nothern glaciated coastal zones with isostatically active coastlines. (The same caution is necessary if we attempt to adopt wider-scale energetically based explanations for the initial exploitation of marine resources in late Pleistocene times in Europe, North Africa, and the North Pacific region. However, the work of Clark et al. (1980) on the north coast of Spain has clearly demonstrated that initial involvement with marine resources in that region occurred under a situation of little or no environmental change.) Optimal foraging theory also helps to explain why lower-calorie, lowfat seafoods may sometimes be dropped from an optimal diet-as, for example, in the famous case of the prehistoric Tasmanians, who stopped eating fish when sea mammals became more abundant under cooler water temperature conditions (Allen 1979; Bowdler 1980; Horton 1979; see also Jones 1977, 1978). Eventually, similar phenomena may help to explain the fish taboos that exist in a number of human societies (Aschmann 1975; Simoons 1974a, 1974b). The purely energetic concerns of optimal foraging theory may, in fact, be more powerful in helping to explain why particular seafoods are utilized than in explaining the use of coastal resources in general. Seafood exploitation can provide net caloric returns in excess of 5,000 kilocalories per hour, and under exceptional circumstances (e. g., the netting of migratory, schooling fish in estuaries during annual runs), it can provide net caloric returns in excess of 18,000 kilocalories per hour (Lischka 1979:5; Perlman 1976: 117), mostly because search times are relatively low. Terrestrial mammals such as deer generally provide net caloric returns of about 3,000 to 5,000 kilocalories per hour (Perlman 1976), but, again, short-term concentrations of aggregated species such as caribou may yield values as high as 27,000 kilocalories per hour (Yesner n. d. a). In fact, if the distinction between coastal and terrestrial resources were truly substantial, it would be difficult to explain why coastal resources in so many parts of North America, for example, were utilized virtually as soon as they became available. Of course, large Pleistocene megafauna may have exceeded the energetic returns of modern terrestrial fauna, at least on a shortterm basis; hypothetically, this could have necessitated the exploitation of "secondary resources," including marine foods, when the megafauna became


11. Life in the "Garden of Eden"

extinct. However, the encounter rates with these species may have been so low as to make these energetic calculations almost meaningless, unless significant food storage was taking place. In addition, per unit energy yield is not the only factor-nutritional or otherwise-that needs to be considered in evaluating the importance of marine diets. Additional factors include those discussed below.

Protein Yields
Osborn (1977) has greatly overemphasized the distinction in protein yields between marine and terrestrial resources. For example, the value that he cites for marine fish flesh (17 percent protein) is close to a minimum value; some marine fish species (such as dogfish and flounder) range as high as 25 percent in protein content, equivalent to or exceeding terrestrial mammals such as caribou, moose, and deer (Arimoto 1962). Osborn's (1977) figure for whalemeat is also fallacious; whalemeat averages about 23.3 percent in protein content, not 10 percent as he suggests. Finally, species of shellfish other than those cited by Osborn (1977) range as high as 24.6 percent in protein content (Borgstrom 1962), and crustaceans such as crab contain up to about 20 percent protein (Arimoto 1962). Furthermore, a great deal of protein is present in shellfish liquids, which are frequently discarded in modern nutritional analyses but would have been important to prehistoric consumers, particularly if the shellfish were eaten raw. There is also evidence to suggest that the digestibility of marine protein is enhanced if eaten raw (Borgstrom 1962). Even certain seaweeds (e. g., eelgrass) harvested by coastal hunter-gatherers contain up to 13 percent protein by weight (Felger and Moser 1973). Beyond simple percentages, marine fish and shellfish yield high-quality proteins, with nitrogen values greater than or equal to that of casein. Both marine fish and shellfish are particularly good sources of lysine, histidine, and arginine. They are somewhat deficient in valine, phenylalanine, and the sulfur-containing amino acids (cystine and methionine), but these are found in greater concentrations in shellfish than in marine fish. It has also been found that specific body parts such as fish eyes or sea mammal flippers-parts that are often considered delicacies-may have particularly favorable amino acid ratios (Bergstrom 1962) and may contain larger amounts of fat (Heller and Scott 1962). Nietschmann (1972, 1973) has demonstrated that a marine diet produces adequate protein to supply coastal populations in Central America, and Meehan (1977) has demonstrated the same for coastal Aborigines in Australia. Experimental evidence from both California (Lischka 1979; Lischka and Sheets 1980) and Alaska (Erlandson 1983) has shown that shellfish collected by a single individual are sufficient to provide for the protein needs of 100 to 500 persons per


IV. Pre-State Foodways

day. In low-latitude regions, the protein supply generated by marine resources is generally sufficient to permit trade with interior-adapted peoples for highcalorie foods such as grains or other plant foods (Stark and Voorhies 1978). Tropical groups in coastal environments are often taller, on average, than neighboring interior groups, and clinical evidence for protein malnutrition among coastal groups is rare (Borgstrom 1962). In fact, the case has been made that among hunter-gatherers in tropical environments, the protein provided by marine foods may have been critical to subsistence in some areas, at least after the extinction of the Pleistocene megafauna and the evolution of the modern biotic zones. One reason for this is the relatively low nitrogen content (and therefore relatively low biological vlaue) of proteins in tropical plants, which use C4 metabolic pathways (Russell-Hunter 1970), as well as the limited secondary productivity of tropical forests. It has also been suggested that horticulture involving low-protein root crops in areas such as Polynesia would have been difficult without the existence of a marine food base (Harris 1982).

Vitamins and Minerals
Seafoods are excellent sources of a number of vitamins, including vitamins A, D, and niacin (Heller and Scott 1962). Vitamin A is particularly plentiful in sea mammal organs, and vitamins A, B 1 , and B2 are concentrated in many fish roes, although they are subject to loss in dried or otherwise stored seafoods (Borgstrom 1962). Mollusks and crustaceans contain large amounts of thiamine, as well as electrolytes such as sodium. Eating whole fish provides a good source of fluorine (up to 250 parts per million), while iodine and Vitamin B 12 are also obtainable from seaweeds (Felger and Moser 1973).

Maximum Sustained Yield
As noted previously (Yesner 1980), migratory marine species can often be intensively exploited with a high maximum sustained yield. In addition, even nonmigratory marine species can withstand culling rates of about 10 to 14 percent, higher than the maximum sustained yield of most terrestrial species (Yesner 1980). It is true, of course, that a good deal of evidence of overexploitation of marine resources has been observed, both ethnographically and archaeologically (Botkin 1980; Shawcross 1975; Swadling 1976, 1977a; 1977b). This seems logical, given that marine hunters tend to be relatively sedentary, exploit relatively large territories, and have relatively high population densities (see "Consequences of the Adoption of Marine Diets" below). Modern marine hunters have certainly been held accountable for the extinction of slow-re-


11. Life in the "Garden of Eden"

producing species such as the northern sea-cow in Alaska (Domning 1972) and captive residents of offshore islands such as the sea-mink in Maine (Loomis 1911). In my own research in Maine, I have found evidence for overexploitation of shellfish, indicated both by size diminution and by a switching back and forth between the species collected (Yesner 1983). Undoubtedly the diminution in shellfish size is the result of a characteristic harvesting pattern of selecting larger individuals (Yesner 1983), a practice guaranteed to affect shellfish during their period of maximum fertility. However, most of the effect of shellfish overexploitation seems to occur within a few hundred years after the initial colonization of a site, suggesting that exploitation patterns stabilize, possibly as groups develop schedules that allow "fallowing" of individual sites for some years while exploitation of alternative sites continues; this is made possible by the more or less ubiquitous distribution of shellfish as well as the mobility permitted by the use of boats. In addition, shellfish populations are extremely difficult to extirpate completely, because small individuals with low net yields are rarely collected at all; thus, populations can almost always regenerate if abandoned for short periods of time. Sedentary sea mammals and sea birds obtained from rookeries on offshore islets can also be subjected to overexploitation, requiring a shift to the exploitation of alternative prey types, as is demonstrated in the archaeological record of the Aleutian Islands (Yesner 1977; Yesner and Aigner 1976). However, these types of resource exploitations are, again, highly localized processes, solutions for which are well within the capabilities of the shifting resource exploitation strategies of individual hunting bands.

Lower Dependency R alios
Because both old people and children are able to engage in activities such as shellfish collection, and because they have lower caloric requirements, they are virtually able to support themselves in coastal zones and do not act as a sump for the population's resources. This may be what has led to the greater life expectancies and more broadly based population pyramids characteristic of regions that depend almost exclusively on marine resources, such as the Aleutian Islands (Laughlin 1972; Yesner 1980). On the basis of the above factors, it should be clear that although coastal resources cannot be considered a Garden of Eden, neither are they entirely second-rate resources as suggested by Osborn (1977). A third model, therefore, needs to be considered-one that offers a more satisfying explanation for the adoption of seafood diets by human societies. This model is based on a feature of those diets not previously discussed: their spatial and temporal



IV. Pre-State Foodways

The Third Model: Risk Reduction, Resource Seasonality, and Energy Flow
Perhaps the greatest contrast between terrestrial and marine foods lies in the much greater spatial and temporal reliability of the latter. Numerous sea mammals, for example, are locally sedentary residents that can be exploited at predictable locations (haulouts and rookeries). Many bottomfish are year-round residents of predictable locations and are thus easily exploited with the proper equipment. Many sea birds have localized colonies at which they can be easily obtained; this is often attested to by the high proportions of immature individuals of these species in the archaeological record. However, sessile invertebrates (e.g., shellfish and crustaceans) are the most "reliable" marine resources of all, both spatially and temporally. Their spatial reliability (predictability) is attested to by the frequently close association of coastal occupation sites with shellfish beds, reefs, and strandflats (Yesner 1980, 1983). Their temporal reliability is due to their relatively low fluctuation in numbers, particularly in areas free from ice scour and not subject to frequent changes in currents or movements of sediment. The resistance of shellfish populations to total depletion, as noted above, is another factor in their temporal reliability; Lischka (1979), for example, has noted that shellfish of "harvestable" size (i. e., over about five centimeters in length) often still exceed densities of 1,200 grams per square meter in areas intensively exploited by humans today. Even migratory marine species are highly reliable in space and time. Migratory sea mammals, birds, and fish (including anadromous fish) follow predictable migration routes at specific times of the year. In the case of sea mammals, lower search costs make the interception of migrating individuals an efficient hunting strategy. On the opposite end of the spectrum, anadromous fish have high search costs along their migratory routes and are therefore generally exploited only at their point of arrival (i. e., spawning streams). Sea birds tend to be intermediate in terms of both search and retrieval costs. The storability of seafoods-primarily through drying or smoking-vastly enhances their reliability. The high salinity of these meats facilitates the storage process, but the high fat content of some marine species often makes it more difficult. Methods therefore tend to depend on fat content; drying is apparently the method of choice for preserving lean fish like bottomfish, whereas smoking (or freezing, where possible) is used to preserve fattier fish such as salmon and other anadromous species. Semi-subterranean storage pits or cache pits for fish, such as the ones recently discovered on the Maine coast (Yesner 1984a), also attest to the important process of cold storage during winter and spring months in Arctic or northern temperate zones. In sum, seafoods tend to be highly reliable resources. Paradoxically, some of the highest-ranked resources in terms of net caloric yield (e.g., anadromous


11. Life in the "Garden of Eden"

fish) are less reliable than lower-ranked resources (e. g., shellfish). This yields a significant "trade-off" potential in coastal zones between highly caloric but less reliable foods and those that are less caloric but more reliable. Hayden (1981) has suggested that the history of cultural evolution has involved continuing efforts by humans to increase the reliability of their subsistence (and thus the security of their existence). Hayden included the initial exploitation of seafoods as a part of that security-seeking pattern. Whether or not his analysis is valid, it is important to note that there is a difference between increasing the reliability of subsistence in a total sense (i. e., through expanding dietary breadth) and seeking out individually reliable food resources (such as seafoods) for exploitation. Utilizing this notion of seafoods as reliable resources, I suggest the following as a realistic scenario for the development of coastal lifeways. First, with the extinction of the "valuable" Pleistocene megafauna, diet breadth expanded at the bottom to encompass a number of resources with lower net yields. During the early Holocene, this primarily meant the exploitation of a wider variety of smaller terrestrial fauna. Then, during mid-Holocene times, as the thermal maximum developed, the width of the temperate zone (which had been compressed under previous glacial conditions) began to expand. This in turn resulted in the expansion of the area characterized by strong resource seasonality. At the same time, these environments were becoming increasingly patchy, thus increasing the exploitation costs for terrestrial resources. Under these conditions marine resources became advantageous, not only because of net caloric yields, fat, and protein content, but also because they were available at the times of the year when terrestrial resources were least productive (in terms of plants) or least accessible (in terms of animals): late winter and spring. It is undoubtedly no accident that seasonal coastal sites throughout the world tend to be occupied during precisely this period (Yesner 1980). It is also noteworthy in this regard that many shellfish species (in North America, oysters and scallops, for example) tend to have a "spring peak" in protein and fat content that appears to be related to the timing of upwelling events (Borgstrom 1962). Thus, the worldwide pattern of spring utilization of coastal resources may be related as much to peaks in the protein and fat content of shellfish as to declines in the availability of terrestrial resources at this time of year. Under these more seasonal conditions, the storability of marine resources would also have made them attractive. A precondition for this possibility was the development of storage technology itself in late Pleistocene and early Holocene times. In short, the wide adoption of coastal resources during mid-Holocene times may have been a function of the increased seasonality of environmental perturbations as well as environmental patchiness. These features are particularly notable in some of the regions where coastal adaptations first developed (e. g., South Africa, the circum-Mediterranean region, Japan). The addition of sea-


IV. Pre-State Foodways

foods to the diet served a function similar to the evolution of storage techniques in that it allowed the evening out of energy flow during the annual cycle, thus minimizing subsistence risk and contributing to population survival. Stabilization of the food supply, including the use of aquatic resources, appears to have been a general goal of late glacial and postglacial hunter-gatherer economies, particularly those in seasonal, patchy, and/or marginal environments such as temperate forested regions (Zvelebil 1986). Why did this phenomenon not occur during earlier (inter-glacial?) periods that may have been characterized by similar circumstances? Ultimately the answer may be found in Binford's (1983) and Cohen's (1977) model of continuous, worldwide population growth during the Pleistocene period. The eventual result of this population growth was greater population packing and corresponding diminution of habitat, which in tum limited the alternatives available when highly ranked resources declined in abundance. The adoption of coastal resources thus allowed an expansion of dietary breadth when increased population density removed the opportunity to survive by continuing to exploit more "valuable" resources. In earlier periods the relatively high-cost resources could be ignored because population "pressure" was not as great. Associated with this transition toward coastal lifeways must have been a significant "retooling" cost in terms of both technology and socioeconomic organization. This would include the costs involved in the development of efficient boat transportation and the complex technology required for harvesting sea mammals and large fish species, as well as the organization of the communal search, pursuit, retrieval, and division of the kill necessitated by the exploitation of those species. It has also been suggested that because many marine species (particularly sea mammals) are largely submerged, their characteristics would have been less discernible to hunting and gathering peoples, thus delaying any automatic transfer of hunting techniques from terrestrial to marine animals. These retooling costs are related to what Binford (1983) has recently termed the "inertia" principle: the idea that once a population adapts to a particular lifeway, it is difficult for it to face the energetic and behavioral demands of adapting to a completely new one. This has certainly been argued for the origins of agriculture, and Binford (1983) implies that it might be applied to the use of the oceans as well. Nevertheless, cases such as the Tanaina Athapaskans of southcentral Alaska demonstrate that maritime adaptation, involving the hunting of at least medium-sized sea-mammals as well as shellfish collection, can be developed among terrestrial hunter-gatherers within a relatively brief period of time (no more than a few hundred years). At the same time, it is important to recognize that, just as with the development of agriculture, the intensification of maritime adaptation often did take place over a considerable period of time, and often proceeded by a series of stages. A common scenario, for example, involves an initial exploitation of sea-


11. Life in the "Garden of Eden"

mammals, followed by an increased exploitation of anadromous fish and/or other large or schooling fishes, and finally a relatively late concentration on shellfish and other marine invertebrates, depending upon the relative availability of these species (Ames 1985). Further intensification of fish exploitation may involve attempts to widen the resource base by obtaining smaller fish, nonschooling fish, or those located further from the coastal zone (often with the use of more complex fishing equipment) and/or it may involve changes in social organization that allow further intensification of anadromous fishing. Further intensification of shellfish exploitation may similarly involve exploitation of harder-to-obtain species, as well as smaller individuals of more "valuable" species-i. e., those that normally provide higher returns for energy invested in exploiting them (Yesner 1984b). Among societies that are buffered by a wide variety of marine resources, intensification of sea-mammal hunting (e. g., the hunting of large baleen whales) is not often observed (Ames 1985). This seems to have occurred primarily in the Arctic zone, beyond the northern limits of both anadromous fish and shellfish populations. In sum, coastal resources do not make up a Garden of Eden, but neither are they clearly second-rate resources in the sense proposed by Osborn (1977). The relatively late adoption of this class of foods is not explained by appealing to such simple models, but instead by developing more complex models relating the use of marine resources to environmental change, technological development, and human population growth. As noted above, the frequent use of coastal resources during the spring season suggests that at least one purpose of utilizing those resources was to minimize subsistence risk at times of the year when other resources were either unavailable or inaccessible; this appears to have occurred in environments with significant seasonal or longerterm fluctuations in the food supply. At the same time, such foods contained important nutrients that might otherwise have been limiting factors in the food supply, even though they may have been relatively low in per unit energy yield. However, because of that energy cost, the exploitation of coastal resources did not occur until population growth had reduced locally available alternatives and forced an expansion of dietary breadth and "retooling" costs had been met. (In keeping with the tenets of optimal foraging theory, at no time was the relative abundance of these resources an issue.) Once the above set of processes is understood, the fact that coastal resources, often highly productive as well as reliable, were not exploited until relatively late should not be considered any more of a paradox than the fact that agriculture, which also greatly increased both the productivity and the security of human subsistence, was even later in its adoption. Perhaps the lack of a clear-cut advantage or disadvantage to the use of marine resources, however, is the reason why, even in those temperate zones where aquatic resources were more prevalent in the diet (Lee 1968), seafoods


IV. Pre-State Foodways

rarely appear to have formed the greater part of that diet. Meehan's (1977, 1982) work suggests that shellfishing, for example, never provided more than about 15 percent of the diet of coastal peoples. The major exceptions here, of course, are high-latitude populations with a high degree of dependence on coastal resources, particularly sea mammals and fish. In many cases the dependence on marine resources was a function not only of the productivity of the coastal zone itself, but also of the productivity, stability, and nutrient yields of interior environments adjacent to the coast. Unfortunately the degree of dependence of prehistoric populations on marine resources can be assessed with confidence only when very detailed archaeological faunal and floral analyses can be undertaken and sampling effects can be sufficiently controlled to allow rigorous quantitative dietary estimates. However, the advent of stable carbon, nitrogen, and strontium isotopic ratio analyses for human bone has allowed us to reconstruct the contribution of marine resources to prehistoric diets with a greater degree of precision. Recently, for example, it has been established that about 90 percent of the protein intake on the British Columbia coast was from marine resources (Chisholm, Nelson, and Schwartz 1983), which exceeds available ethnographic estimates (Murdock 1967), as well as archaeological estimates (Boucher 1976), by about 35 percent. These discrepancies are worth noting, particularly for areas in which little or no quantitative ethnographic data exist. For example, I have recently undertaken analyses of both stable carbon and nitrogen isotopes from a series of skeletons from Moshier Island in Casco Bay, Maine (dated to ca. 1000 A. D.), which suggest that 80 percent of the diet was from marine resources; this appears to exceed the estimates from faunal analysis by about 20 percent. However, such statistics may be meaningless unless we know how much seasonal movement and food storage was undertaken by these groups. In the case of our research in coastal Maine, fish vertebra and shellfish sectioning has demonstrated a pattern of seasonal transhumance; the coast was occupied primarily during the spring (i. e., from late winter to early summer), and codfish were stored in the large pits found in some coastal sites. Unfortunately, since the preservation of faunal remains in interior sites is very poor, it is impossible to fill out the seasonal round in the same rigorous fashion. It is doubtful, therefore, whether ecofactual analyses from coastal sites will ever be as rigorous as the chemical analyses now being developed for the examination of human bone.

Consequences of the Adoption of Marine Diets
To this point I have focused primarily on the causes, rather than the consequences, of the adoption of marine diets by human societies. In this section, I want to tum instead to four related aspects of human society affected by the


11. Life in the "Garden of Eden"

shift to seafood diets: (1) population mobility; (2) population density; (3) population health; and (4) sociopolitical complexity. Let me deal with each of these in turn.

Population Mobility
Elsewhere (Yesner 1980) I have characterized coastal lifeways as "semi-sedentary," typified by "communities whose members shift from one to another fixed settlement at different seasons, or who occupy more or less permanently a single settlement from which a substantial proportion of the population departs seasonally to occupy shifting camps" (Binford 1980: 13). Coastal settlement allows "central-place foraging," primarily with the use of boats, a strategy that promotes sedentism and thereby reduces mobility costs in the exploitation of resources. However, a great deal of variation in settlement pattern among coastal societies is encompassed within the general framework of "semi-sedentary," central-place foraging. For example, on one extreme the Yahgan of southern Patagonia used canoes for daily outings but returned to different camps every few days, possibly to prevent overexploitation of shellfish beds in an area where marine invertebrates were rather scantily distributed, sea mammal resources were limited and anadramous fish resources were absent (Stuart 1972, 1977). On the other extreme, Northwest Coast Indians and Bering Sea Eskimos were characterized by relatively little mobility, primarily of task groups and nuclear family units; such sedentism seems to be related in particular to the availability of anadromous fish (Schalk 1977). Among the factors determining the degree of mobility necessary to obtain marine resources are the following: 1. The diversity and abundance of resources available within the coastal zone 2. The productivity of terrestrial environments adjacent to the coastal zone 3. Temporal variations in the strength of nutrient upwelling systems 4. The frequency and timing of migrations of marine species, and the importance of these migratory species in the overall diet 5. Variations in marine resource aggregation (e.g., the number of rookeries, colonies, or haulouts, and their spatial distribution) The nutritional quality of coastal resources also has important constraining effects on population mobility. Several years ago I argued (Yesner 1979) that mesolithic coastal populations that focused almost exclusively on shellfish would have been forced into some degree of transhumance in order to supply some important nutrients that would otherwise be missing in the diet. It is equally important to consider fluctuations in the nutritional quality of seafoods, something that affects nearly all coastal populations regardless of diet, but


IV. Pre-State Foodways

again particularly affects those dependent on marine resources. The fat content of sea mammals and sea birds fluctuates seasonally, like that of terrestrial mammals and birds. Among anadromous fish and shellfish, fluctuation in fat content is related primarily to spawning or reproductive cycles, which vary among species. As noted above, shellfish also vary in protein content, having a "protein peak" that, again, differs by species. These variations in protein and fat content could well be limiting factors on coastal populations and need to be taken into account more often in developing models of seasonal population movements by these peoples. An additional factor to be considered in developing such models is the timing of "red tide" infestations in shellfish and other seafoods during the late spring and early summer.

Population Density
The constraints of coastal lifeways on human population densities are more difficult to assess. High local population densities among coastal groups may in some cases be less a function of ecosystem richness per se than of local concentrations of resources ("species packing") in areas such as estuaries, island passes (where upwelling takes place), peninsulas (which funnel migratory seamammals), or areas where streams, lakes, or marshes about the coast (where there is a high diversity of fish, birds, and small mammals). In addition, the productivity of the terrestrial zone adjacent to the coast may have as much to do with aboriginal coastal population density as the richness of the coast itself, particularly in lower latitudes where the major function of coastal resources may be as a dietary protein supplement (Osborn 1980). On the basis of both the archaeological and ethnographic records, it has been argued that population densities are higher in coastal regions. Numerous authors have argued that increased sedentism alone offers a robust explanation for population growth, resulting from increased fertility associated with changes in exercise patterns, male absenteeism, female work load (including carrying of infants), and desire for children. If population growth potential is indeed higher for coastal societies, this suggests that pressure on resources would be likely to occur every so often. Seeking objective evidence for such population pressure, however, is difficult at best. If we follow Cohen's (1975) caveat, however, that one of the best dicators of population pressure is the increasing use of more marginal habitats, then my research in both the Aleutian Islands and on the coast of Maine has demonstrated that, indeed, continued population growth and pressure on resources have been occurring since mid-Holocene times. This expansion involves the use of microenvironments with less species diversity, such as


11. Life in the "Garden of Eden"

straighter, less complex (and hence less biotically diverse) coastlines, as well as areas with lower species abundance, such as small, offshore islands (Yesner 1984b; Yesner and Aigner 1976). Even for North America as a whole, it is possible to show that areas such as the California Channel Islands or the Florida Keys were occupied relatively late in the prehistoric record, at the end of a period of continuous local population growth.

Population Health
'[he diets of coastal peoples generally provided enough calories to support relatively large population densities (Yesner 1980) and may be related to an earlier menarche as well as greater life expectancy among coastal societies. This does not mean, of course, that no health problems were created by occasional population "pressure" on resources or by inadequacies in the marine diets themselves. There is, however, very little evidence of nutritionally related disease among coastal groups. For example, the only major forms of paleopathology consistently associated with coastal peoples are activity-related ones, particularly dental attrition (and its consequences), and degenerative and traumatic disorders (e. g., osteoarthritis). These are found with particularly high frequency in Arctic populations but are also high in other coastal populations, such as aboriginal California Ourmain 1983; Leigh 1928; Walker 1978). We have found a similar pattern in our skeletal series from Moshier Island on the Maine coast (Yesner n. d. b). In areas in which skeletal indicators of acute nutritional deficiency (such as Harris lines) have been found, as in coastal California (McHenry 1968), they have been associated with periods in which less use of marine foods was taking place. However, it is still possible that chronic nutritional stress may have been occurring in these populations. The fact that the frequency of enamel hypoplasia (16.7 percent) is somewhat higher than that of Harris lines (10.4 percent) in these coastal California groups may indeed suggest some nutritional stress (McHenry and Schulz 1975; but see Chapter 10). Additionally, cribra orbitalia has been noted among groups such as California coastal Indians (Walker 1986) and Northwest Coast Indians (Cybulski 1978; Gordon 1974), but "it is difficult to reconcile the findings in terms of possible deficiency in dietary intake of iron in the earlier British Columbia population," since the aboriginal diet of fish, shellfish, and sea mammals was iron-rich (Cybulski 1978:37). Other factors, such as hereditary patterns or infectious diseases, might be at fault, and infectious diseases or metabolic disorders might also explain some of the "growth-arrest" indicators seen in human skeletons from coastal areas. In this case, of course, an increased load of infectious disease might reflect the effects of higher population density, although no evidence has been found of infectious


IV. Pre-State Foodways

disorders associated with iron-deficiency anemias (e. g., tuberculosis), which do leave their mark on bone.

Sociopolitical Complexity
As noted earlier, the range of cultural complexity associated with maritime populations ranges from band-level organization (e. g., among the Yahgan) to incipient state-level societies (e. g., on the prehistoric coast of Peru). However, these extremes appear to be associated with an unusual paucity and abundance of resources, respectively Oackson and Popper 1980; Osborn 1977; Stuart 1972). The most commonly cited models for coastal societies involve "bigman" types of political organization or chiefdoms characterized by incipient social stratification. A greater degree of cultural complexity seems likely to occur where anadromous fish are a significant element in the diet of coastal peoples, since their predictability in space and time allows for a good deal of control over these resources, including the restriction of access to the more productive fishing locales (Schalk 1977). Hierarchical leadership may have evolved in such regions in order to coordinate the complex scheduling required for catching, processing, and storing spatiotemporally aggregated anadromous fish resources (Ames 1985). In general, coastal groups tend to display a greater degree of sociocultural complexity than other hunter-gatherers, and Cohen (1981) has suggested that this may directly result from the higher population density achieved in coastal regions. He cites the following connections between high population density and increased cultural complexity: (1) less room for spatial adjustment to resource failure; (2) fewer "emergency" food resources available; and (3) less "social flexibility" in adjustment to resource variations. Thus, "the elaboration of material culture and of social hierarchies . . . may reflect a social response to economic vulnerability in the face of resource failure-the vulnerability resulting from high population density" (Cohen 1981:275). Two commonly employed archaeological indices of greater cultural complexity found among coastal peoples are the elaboration of grave goods and the presence of exotic items, reflecting an elaboration of ritual and the development of widespread trade networks managed by an elite group. On the basis of these indices, King (1978) and others have demonstrated that a cultural complexity above the band level existed in aboriginal California. Zvelebil (1986) has developed similar arguments for Mesolithic Europe. In northern New England as well, the widespread appearance of exotic materials such as Lake Superior copper in coastal burials suggests some evolution of cultural complexity. (In all three cases however, there is relatively little difference in the quantity or quality of grave goods buried with different individuals, suggesting the presence of


11. Life in the "Garden of Eden"

only incipient social stratification.) These archaeological data from coastal societies help us to make sense out of ethnographic accounts of, for example, the so-called sagamores and sachems on the New England coast, and the doyons of the Aleutian Islands. Although marine resources may have most commonly supported only incipient stratified societies, there is evidence to suggest that they may have helped to support more complex societies as well. Lange (1971), for example, has made an excellent case for the importance of marine resources in supporting the framework of the lowland Maya civilization, and, as suggested earlier, these resources were also important to highland Meso-American civilizations (Stark and Vorrhies 1978). Coastal Peru, however, still appears to be unique in the degree of cultural complexity that was supported by a marine food base there (Moseley 1975; Osborn 1977). In that area, original interest in the coast may have been fed by a tremendous local seasonal variation in terrestrial food supply, as well as by the availability of a rich coastal zone. An unusual degree of fluctuation in coastal resources (which, as noted above, are generally more reliable over time), due to the phenomenon of El Nino, in turn promoted increased cultural complexity in Peru. In addition, the juxtaposition of a relatively rich coastal zone with a depauperate interior zone created unique ecological conditions favorable to the evolution of redistributive systems, which in turn contributed to cultural complexity in that area; Kelly (1983) has recently offered a similar argument for the origin of the Northwest Coast cultural pattern. Population "spillover" from densely populated coastal zones to adjacent resource-poor interior zones may also have promoted an increased interest in horticulture, whether in Peru or in the U. S. Southwest (Glassow 1972). Thus, to return to Binford's (1983) original question about the apparent differences in receptivity to agriculture in coastal regions, it may well be that the reasons that California Indians did not adopt agriculture had more to do with the relative lichness of both the coast and the (acorn-using) interior zone than with the richness of the coast alone. One final element needs to be considered as well: that of resource failure. Zvelebil (1986: 114) has recently noted that "farming was not necessarily advantageous . . . for communities that specialized in exploiting aquatic resources," and that this may have led to a delay in commitment to a farming way of life among such littoral hunter-gatherers in temporate Europe. Yet, paradoxically, it was the commitment to such a coastallifeway that led these mesolithic .Europeans down the road toward agriculture. Although their lifeway "was capable of supporting relatively high population densities and of absorbing some fluctuations in resources," when major crises occurred in sea-mammals and a shellfish in various parts of Scandinavia-whether as a result of climate change or culturally-induced factors-such coastal foraging economies could no longer be maintained, and were quickly replaced by farming populations. In this light,


IV. Pre-State Foodways

the "Garden of Eden" principle can be seen both as the reason for the original rejection of agriculture and as the eventual causative agent for the acceptance of agriculture. Risk-reducing strategies are apparently maintained by human societies until the risks themselves make them no longer tenable; at that point, other risk-reducing strategies are implemented. Certainly, like the maritime adaptations that preceded them, the ostensibly risk-reducing farming strategy contained within it the seeds of even greater risks for growing human populations.

Coastal resources are neither a "Garden of Eden" nor second-rate resources; they allow neither total mobility nor permanent settlement; and a wide variation in degree of social stratification is found among societies exploiting these resources. Variations in these phenomena may have more to do with the relationship of coastal and interior zones than with the nature of coastal resources alone. However, while recognizing this variability in subsistence, settlement, and sociopolitical organization, we can discern a number of consistent patterns in the preceding analysis of coastal adaptations. To begin with, the pattern of delayed adoption of marine diets is observable worldwide. This phenomenon can best be explained as a result of concurrent forces of environmental change (especially increased seasonality and environmental patchiness) and population growth, during a period of rapid technological change, including the development of food storage technology. The consequences of the adoption of marine diets included more sedentary lifestyles and further increases in population density; these increases may have had some consequences for individual health as well as for the generation of social complexity. Our work on the Maine coast is also beginning to suggest that, in a feedback fashion, the population growth associated with semi-sedentary coastal societies forces an even greater commitment to a coastal diet. As with the origins of agriculture, an element of irreversibility seems to be associated with this important transition in human lifeways.

Much of the fieldwork in the Aleutian Islands and the Maine coast was supported by grants from the National Science Foundation and the Maine Historic Preservation Commission. I would like to thank all of the conference participants who commented on my


11. Life in the "Garden of Eden"

paper, and particularly Jane Buikstra, who suggested revisions that were extremely helpful in the redrafting of some of the ideas.

References Cited
Allen, H. 1979 Ames, K. 1985 Left Out in the Cold: Why the Tasmanians Stopped Eating Fish. Artefact 4: 1-10. Hierarchies, Stress, and Logistical Strategies Among Hunter-Gatherers in Northwestern North America. In Prehistoric Hunter-Gatherers: The Emergence of Cultural Complexity, T. D. Price and]. A. Brown, eds., pp. 155-80. Orlando: Academic Press. The Role of Fish in the Japanese Diet. In Fish as Food, G. Borgstrom, ed., pp. 370-71. New York: Academic Press.

Arimoto, K. 1962

Aschmann, H. Culturally Determined Recognition of Food Resources in the Coastal 1975 Zone. Geoscience and Man 12:43-47. Binford, L. R. 1980 Willow Smoke and Dogs' Tails: Hunter-Gatherer Settlement Systems and Archaeological Site Formation. American Antiquity 45: 4-20. 1983 In Pursuit of the Past. New York: Thames and Hudson. Borgstrom, G. 1962 Shellfish Protein-Nutritive Aspects. In Fish as Food, G. Borgstrom, ed., pp. 115-47. New York: Academic Press. Botkin, S. 1980 Effects of Human Exploitation on Shellfish Populations at Malibu Creek, California. In Modeling Change in Prehistoric Subsistence Economies, T. K. Earle and A. L. Christenson, eds., pp. 121-40. New York: Academic Press. Boucher, N. D. 1976 Prehistoric Subsistence at the Helen Point Site. M.A. thesis, Simon Fraser University, Burnaby, British Columbia. Bowdler, S. 1980 Fish and Culture: A Tasmanian Polemic. Mankind 12:334-40. Chisholm, B. S.; D. E. Nelson; and H. P. Schwartz 1983 Determination of the Relative Intakes of Marine and Terrestrial Protein by the Prehistoric Inhabitants of the British Columbia Coast. Current Anthropology 24:396-98. Clark, G.; L. Straus; ]. Altuna; and]. Ortea 1980 Ice Age Subsistence in Northern Spain. Scientific American 242 Gune): 142-53.


IV. Pre-State Foodways

Cohen, M. N. 1975 Archaeological Evidence for Population Pressure in Pre-Agricultural Societies. American Antiquity 40: 471-75. 1977 The Food Crisis in Prehistory. New Haven: Yale University Press. 1981 Pacific Coast Foragers: Affluent or Overcrowded? In Affluent Foragers: Pacific Coasts East and West, S. Koyama and D. H. Thomas, eds., pp. 275-95. Osaka: National Museum of Ethnology. Cybulski, ]. 1978 Cribra orbitalia, a Possible Sign of Anemia in Early Native Populations of the British Columbia Coast. American Journal of Physical Anthropology 47:31-40. Domning, D. P. 1972 Steller's Sea Cow and the Origins of North Pacific Aboriginal Whaling. Syesis 5: 187-89. Erlandson, ]. 1983 The Role of Shellfish in Coastal Economies: A Protein Perspective. Paper read at the meeting of the Alaska Anthropological Association, Anchorage. Felger, R., and M. B. Moser 1973 Eelgrass (Zostera marina) in the Gulf of California: Discovery of Its Nutritional Value by the Seri Indians. Science 181:355-56. Fischer, S., and P. C. Weber 1984 Prostaglandin I] Is Fonned in vivo in Man After Dietary Eicosapentaenoic Acid. Nature 307: 165-68. Glassow, M. 1972 Changes in the adaptations of Southwestern Basketmakers: A Systems Perspective. In Contemporary Archaeology, M. P. Leone, ed., pp. 289302. Carbondale: Southern Illinois University Press. Gordon, M. E. 1974 A Qualitative Analysis of Human Skeletal Remains from Gabriola Island, British Columbia. M.A. thesis, University of Calgary. Harris, D. R. The Prehistory of Human Subsistence: A Speculative Outline. In Food, 1982

Nutrition, and Evolution: Food as an Environmental Factor in the Genesis of Human Variability, D. N. Walcher and N. Kretchmer, eds., pp.
15-35. New York: Masson.
Hayden, B. 1981 Research and Development in the Stone Age: Technological Transitions Among Hunter-Gatherers. Current Anthropology 22: 519-48. Heller, C. A., and E. M. Scott 1962 rrhe Alaska Dietary Survey. Washington, D. C.: Government Printing Office. Hill, K. 1982 Hunting and Human Evolution. Journal ofHuman Evolution 11: 521-44. Horton, D. R. 1979 Tasmanian Adaptation. Mankind 12:28-34.


11. Life in the "Garden of Eden"

Jackson, H. E., and V. Popper 1980 Coastal Hunter-Gatherers: The Yahgan of Tierra del Fuego. In The Ar-

chaeological Correlates of Hunter-Gatherer Societies: Studies from the Ethnographic Record, F. E. Smiley et al., eds., pp. 40-61. Ann Arbor:
University of Michigan Museum of Anthropology. Jones, R. 1977 The Tasmanian Paradox. In Stone Tools as Cultural Markers, R. V. S. Wright, ed., pp. 189-204. Canberra: Australian Institute for Aboriginal Studies. Why Did the Tasmanians Stop Eating Fish? In Explorations in Ethnoarchaeology, R. Gould, ed., pp. 11-47. Albuquerque: University of New Mexico Press.


Jurmain, R. D. Paleoepidemiology of a Native Californian Skeletal Population. American 1982 Journal ofPhysical Anthropology 51:211-12. Kelly, R. L. Hunter-Gatherer Mobility Strategies. Journal of Anthropological Re1983 search 39: 277-306. King, T. F. Don't That Beat the Band? Nonegalitarian Political Organization in Pre1978 historic Central California. In Social Archaeology: Beyond Subsistence and Dating, C. Redman et al., eds., pp. 225-48. New York: Academic Press. Lange, F. W. 1971 Marine Resources: A Viable Subsistence Alternative for the Prehistoric Lowland Maya. American Anthropologist 73: 619-39. Laughlin, W. S. 1972 Ecology and Population Structure in the Arctic. In The Structure of Human Populations, G. A. Harrison and A. ]. Boyce, eds., pp. 379-92. Oxford: Clarendon Press. Lee, R. B. 1968 What Hunters Do for a Living; or, How to Make Out on Scarce Resources. In Man the Hunter, R. B. Lee and I. DeVore, eds., pp. 30-48. Chicago: Aldine. Leigh, R. W. 1928 Dental Pathology of Aboriginal California. University of California Publications in A merican A rchaeology and Ethnography 23: 399-423. Lischka, ]. ]. 1979 On the Evaluation of Marine and Terrestrial Food Resources. Manuscript. Department of Anthropology, University of Colorado, Boulder. Lischka, ]. ]., and P. D. Sheets 1980 Comments on "Maritime Hunter-Gatherers: Ecology and Prehistory." Current Anthropology 21: 739-40. Loomis, F. 1911 A New Mink from the Shell-Heaps of Maine. Arnerican Journal of Science 181: 227-29.


IV. Pre-State Foodways

Lumley, H. de 1969 A Paleolithic Camp at Nice. Scientific American 220 (May):42-50. McHenry, H. 1968 Transverse Lines in Long Bones of Prehistoric California Indians. American Journal ofPhysical Anthropology 29: 1-18. McHenry, H. M., and P. D. Schulz 1975 The Association Between Harris Lines and Enamel Hypoplasia in Prehistoric California Indians. American Journal of Physical Anthropology 44:507-12. Meehan, B. 1977 Man Does Not Live by Calories Alone: The Role of Shellfish in a Coastal Cuisine. In Sunda and Sahul: Prehistoric Studies in Southeast Asia, ]. ]. Allen et al., eds., pp. 493-531. New York: Academic Press. 1982 Shell Bed to Shell Midden. Canberra: Australian Institute of Aboriginal Studies. Meighan, C. Maritime Adaptation and Midden Analysis in California. Paper read at 1983 the meeting of the Society for American Archaeology, Pittsburgh, Pa. Moseley, M. 1975 The Maritime Foundations of Andean Civilization. Menlo Park, Calif.: Cummings. Murdock, G. P. 1967 Ethnographic Atlas. Pittsburgh: University of Pittsburgh Press. Nietschmann, B. 1972 Hunting and Fishing Focus Among the Miskito Indians, Eastern Nicaragua. Human Ecology 1:41-67. 1973 Between Land and Water. New York: Seminar Press. O'Connell, ]. F., and K. Hawkes 1981 Alyawara Plant Use and Optimal Foraging Theory. In Hunter-Gatherer Foraging STrategies: Ethnographic and A rchaeological Analyses, B. Winterhalder and E. A. Smith, eds., pp. 99-125. Chicago: University of Chicago Press. Osborn, A.]. Strandloopers, Mermaids, and Other Fairy Tales: Ecological Determi1977 nants of Marine Resource Utilization. In For Theory Building in Archae-

ology: Essays on Faunal Remains, Aquatic Resources, Spatial Analysis, and Systemic Modeling, L. R. Binford, ed., pp. 157-205. New York:
1980 Academic Press. Aboriginal Coastal Population Density: Toward the Resolution of a Paradox. Paper read at the meeting of the Society for American Archaeology, Philadelphia, Pa.

Perlman, S. 1976

Optimum Diet Models and Prehistoric Hunter-Gatherers: A Test on Martha's Vineyard. Ann Arbor: University Microfilms. Russell-Hunter, W. D. 1970 Aquatic Productivity. New York: Macmillan.


11. Life in the "Garden of Eden"

Schalk, R. F. 1977

The Structure of an Anadromous Fish Resource. In For Theory Building

in Archaeology: Essays on Faunal Remains, Aquatic Resources, Spatial Analysis, and Systemic Modeling, L. R. Binford, ed., pp. 207-50. New
York: Academic Press. Shawcross, W. F. 1975 Some Studies of the Influences of Prehistoric Human Predation on Marine Animal Population Dynamics. In Maritime Adaptations of the Pacific, R. W. Casteel and G. I. Quimby, eds., pp. 39-66. The Hague: Mouton. Shodell, M. The Prostaglandin Connection. Science 83:78-82. 1983 Simms, S. R. The Evolution of Hunter-Gatherer Foraging Strategies. Paper read at 1983 the meeting of the Society for American Archaeology, Pittsburgh, Pa. Simoons, F. ]. 1974a Fish as Forbidden Food. Ecology ofFood and Nutrition 3: 185-201. 1974b Rejection of Fish as Human Food in Africa: A Problem in History and Ecology. Ecology ofFood and Nutrition 3: 158-64. Singer, R., and j. Wymer 1982 The Middle Stone Age at Klasies River Mouth in South Africa. Chicago: University of Chicago Press. Stark, B. L., and B. Voorhies 1978 Prehistoric Coastal Adaptations. New York: Academic Press. Stuart, D. E. 1972 Band Structure and Ecological Variability: The Ona and Yahgan of Tierra del Fuego. Ann Arbor: University Microfilms. Seasonal Phases in Ona Subsistence, Territorial Distribution, and Orga1977 nization: Implications for the Archaeological Record. In For Theory

Building in Archaeology: Essays on Faunal Remains, Aquatic Resources, Spatial Analysis, and Systemic Modeling, L. R. Binford, ed., pp. 25186. New York: Academic Press. Swadling, P. 1976 1977a 1977b Changes Induced by Human Exploitation in Prehistoric Shellfish Populations. Mankind 10: 156-62. Implications of Shellfish Exploitation for New Zealand Prehistory. Mankind 11: 11-18. Depletion of Shellfish in the Traditional Gathering Beds of Pari. In The Melanesian Environment, ]. H. Winslow, ed., pp. 182-87. Canberra: Australian National University Press.

l'artaglia, L. ]. 1976 Prehistoric Maritime Adaptations in Southern California. Ann Arbor: University Microfilms. loyama, Y., and T. Kaneda 1962 Nutritive Aspects of Fish Oils. In Fish as Food, G. Borastrom, ed., pp. 149-73. New York: Academic Press.


IV. Pre-State Foodways

Volman, T. P. 1978 Walker, P. L. 1978 1986

Early Archaeological Evidence for Shellfish Collecting. Science 201:91113. A Quantitative Analysis of Dental Attrition Rates in the Santa Barbara Channel Area. American Journal ofPhysical Anthropology 48: 101-6. Porotic Hyperostosis in a Marine-Dependent California Indian Population. American Journal ofPhysical Anthropology 69:345-54.

Yesner, D. R. 1974 Population Pressure in Coastal Environments: an Archaeological 1est. World Archaeology 16: 108-27. 1977 Resource Diversity and Population Stability Among Hunter-Gatherers. Western Canadian Journal ofAnthropology 7: 18-59. 1979 Nutrition and Cultural Evolution: Patterns in Prehistory. In Nutritional Anthropology, N. Jerome et aI., eds., pp. 85-116. Pleasantville, New York: Redgrave. 1980 Maritime Hunter-Gatherers: Ecology and Prehistory. Current Anthropology 21: 727-50. 1981 Archaeological Applications of Optimal Foraging Theory: Harvesting Strategies of Aleut Hunter-Gatherers. In Hunter-Gatherer Foraging Strategies: Ethnographic and Archaeological Analyses, B. Winterhalder and E. A. Smith, eds., pp. 148-70. Chicago: University of Chicago Press. 1983 On Explaining Changes in Prehistoric Coastal Economies: 1'he View from Casco Bay. In The Evolution ofMaritirne Cultures on the Northeast and Northwest Coasts of Alnerica, . ]. Nash, ed., pp. 77-90. Burnaby, B. C., Department of Archaeology, Simon Fraser University. 1984a The Structure and .Function of Prehistoric Households in Northern New England. Man in the Northeast 28:51-72. n.d.a Comparative Hunting Efficiencies for Arctic and Subarctic Species. Manuscript. Portland: Department of Geography-Anthropology, University of Southern Maine. n. d. b Prehistoric Maritime Adaptation in Southwestern Maine: Isotopic Evidence from the Moshier Island Burial Site. Paper presented to Chacmool Conference, Calgary, 1986. Yesner, D. R., and]. S. Aigner 1976 Comparative Biomass Estimates and Prehistoric Cultural Ecology of the Southwest Umnak Region, Aleutian Islands. Arctic Anthropology 13: 91112. Zvelebil, M. Postglacial Foraging in the Forests of Europe. Scientific American 254 1986 (5): 104-15.



The Analysis of Hunter-Gatherer Diets: Stalking an Optimal Foraging Model
MY TIT'LE ALLUDES TO HUNTING. THE IMAGERY IS DELIBERATE: I intend a somewhat predatory engagement with the subject. How do ecological factors affect hunter-gatherer decisions about the harvest of non-produced food resources? What are the ecological strategies of the food quest? And, especially" how does one go about asking these questions? What is a productive or heuristically useful research procedure? I will suggest that the analytic process is a little like foraging itself. The inquiry is, in a sense, the quarry. This paper has three parts: (1) a brief statement about the methodology of evolutionary ecology that focuses on assumptions and attempts to convey a particular attitude toward this kind of invesgiation; (2) a quick summary of applications, giving evidence that the assumptions are not so unrealistic nor the effort so unrewarding as they may seem after the first part; and (3) an exploration that attempts to extend the models while showing more precisely what they cannot do.

A Brief Historical Digression
1 will begin indirectly, with an older but similar biological model, the logistic equation, developed first by Verhulst in the 1840s but promoted and popularized by Pearl in the 1920s. The history of the equation (Kingsland 1982) contains lessons about the strengths and limitations of foraging models. The logistic is a systematic expression of a set of biological relationships (Figure 12.1). It has an easily visualized graphic form, the s-curve. It can be stated as an intuitively appealing equation of several variables. It also raises some hard, though usually implicit, questions. What kind of scientific status do we give to a creature like this? For what is it best used?


IV. Pre-State Foodways

FIGURE 12.1. The Logistic Curve and Equation

([[ == rN


(K-N) K)


NOTE: See Hutchinson (1978) for a definition of the terms in the equation.

Verhulst was after a calculating device (" logistique") derived from first principles. He had practical problems in mind-the limitations of Belgian farmland, for one. In contrast, Pearl developed the formula directly from empirical data and promoted it as a Law, comparable "in a modest way" to Boyle's and Kepler's laws. Pearl devoted much of his scientific work to showing that populationgrowth data sets fit the logistic curve. This required second-order corrections in the equation and some creative attention to the data as well. Pearl's promulgation of the equation in this manner led to clamorous debates, as economists, statisticians, and demographers pointed out that its assumptions were artificial and that the fudging required to preserve its lawfulness was counterproductive. Others approached the equation differently. Lotka realized the potential of closely investigating the r term. His pursuit of that component led to formulas relating population growth rates to age structure. The logistic gave Lotka entry to demographic questions going beyond the equation itself. He said of it: "An empirical formula is . . . not so much the solution of a problem as the challenge to such a solution. It is a point of interrogation, an animated question mark" (quoted in Kingsland 1982:42). Gause (1934) realized that the growthresistance term-(K - N)/K-expressed competitive processes like those envisioned in Darwinian natural selection. He extended the model to two competing species and used it as a tool for the analysis of resource competition and co-existence. In his experimental work he attempted to estimate model param-


12. The Analysis of Hunter-Gatherer Diets

eters directly, but he did not confuse the equation with a realistic portrayal of nature. The debates provoked by Pearl are forgotten; the approaches developed by Lotka and Gause permeate evolutionary biology. Kingsland's summary indicates why:
The logistic curve cannot be tested by comparison with observations, as one would test a scientific hypothesis, for it is neither a law nor a hypothesis, but a logical argument based on a variety of assumptions. By looking at deviations from the logistic curve, however, one can refine these assumptions to gain a more accurate understanding of how a population behaves. The logistic can therefore be useful as a tool of research even though it is not a realistic description of growth. (Kingsland 1982:41)

Optimal foraging models are valuable in much the same way-not as lawful statements about reality, but as structured forms of inquiry, more interesting to stalk than to live by.

Assumptions, Procedures, and Two Basic Models

Optimization as a Working Principle
Optimization is the outcome of selection in a finite environment. Resource competition gives the edge to organisms more efficient at gaining energy and nutrients to expend on more foraging, the search for mates, or avoidance of predators and hazards. This is a neo-Darwinian article of faith, although competition is remarkably difficult to observe in nature (Connell 1975; Schoener 1982; Wiens 1977). The extensive use in evolutionary ecology of optimization and cost-benefit assessments has led to a serious debate about their validity and merits (Gould and Lewontin 1979; Lewontin 1978; Maynard Smith 1978). I argue (Winterhalder 1981a: 15) that optimization principles provide a structured, contingent, and partial guide: an entry to inquiry rather than an acceptable proposition about nature. They are structured because they allow one to devise hypotheses or predictions about adaptive behaviors in well-defined conditions. They are contingent because no set of conditions has complete theoretical generality. They are partial because they assist one in speculating about selection at the expense of attention to other aspects of the evolutionary process. In general, they make explicit and thus open to scrutiny a widespread form of functionalist reasoning (Smith and Winterhalder 1981). The optimization principle relevant to what follows is this: human foragers


IV. Pre-State Foodways

will adopt behaviors that allow them to achieve the highest possible net rate of energy capture while foraging. Note the phrase "while foraging." It circumvents the assumption that hunter-gatherers attempt to maximize (or minimize) resource use in general. It leaves open the question of how long an organism forages. It is a postulate of scarcity for analytic purposes, not a statement about actual scarcity. It is a formal, operational way of stating a commonsense hypothesis: hunter-gatherers will have developed behaviors that make them as skillful and successful as is possible in the capture of game or harvesting of plants, relative to their effort. I will return below to the question of how long they forage. The expectation of optimal results is more heuristic than realistic. Natural design highlights the efficacy of selection, but a careful look often shows that nature is "tinkered" together, to use Jacob's (1977) metaphor. Constraints arise from history, chance, and competing goals (Cody 1974). The appropriate pre-adaptations or phenotypic variance may not be present. Environmental fluctuations may impede or reverse directional evolution. Or the optimal responses for concurrent goals may conflict. Nature is less facile, it is messier, and it is not so single-minded as the methods of the evolutionary ecologist. Thus, the constraints themselves may be "more interesting and more important in delimiting pathways of change than the selective force that may mediate change when it occurs" (Gould and Lewontin 1979:581). Optimization analysis sensitive to this possibility is the best way to uncover the effects of such constraints. The issue of constraints raises one of mechanism. Although Darwinian in origin, optimal foraging models can predict the effects of the selection of behavioral variants not associated with genetic variance (see discussion in Durham 1976; Orlove 1980; Pulliam 1981; Pyke, Pulliam, and Charnov 1977; Richerson 1977).

Simple and Compound Currencies
An optimization assumption requires a currency. Here again there is debate. A simple energy currency is used widely. It addresses several possibilities (Smith 1979): (1) energy may be periodically or chronically scarce; (2) foraging may expose the hunter-gatherer to greater-than-average hazards; or (3) foraging may divert valuable time from other activities. Energy, unlike some nutrients, is an immediate, recognizable need and hence a likely proximate guide to foraging behavior. If, however, resources have several significant nutritional or non-food attributes, or complementary effects, then compound or linear programming models are required. Resources assessed by multiple attributes are not easily


12. The Analysis of Hunter-Gatherer Diets

ranked; those with complementary effects cannot be assigned a value independently of one another (Rapport 1980; Westoby 1978). Few researchers have asked in what circumstances simple or compound currency models are more appropriate. In general, the foods of carnivores are of a high quality and roughly comparable in their nutrient mix. Attaining sufficient energy is likely to subsume other requirements (but see Speth and Spielmann 1983). The mixed diet of an omnivore may be nutritionally adequate even if "chosen" on the basis of an energy currency, but here nutrient considerations are more salient. The low and uneven quality of plant foods makes a presumptive case for compound currencies when these predominate in the diet. Compound currencies require linear programming models, but I am wary of the view that complex nutritional models are more desirable because they appear to be more realistic. They incorporate more elements, but whether they accurately take account of them is more difficult to decide. Few of the input variables are known precisely (see Chapter 6), and although detailed, results can quickly become sensitive to these poorly understood parameters. Ethnographic study Oohnson and Behrens 1982) has shown that specific auxiliary information is needed to "calibrate" a linear programming model to particular circumstances and questions. Simple models are hedged with analytically penetrable uncertainties. As structured questions, their virtues diminish as their structure escapes comprehension. Whatever the currency, most foraging models are deterministic. They are based on average or "expected" values and do not allow for the possibility that decisions respond to stochastic variability in the factors affecting foraging. I will return below to the question of stochasticity and risk (see Jochim 1982).

Models as Creatures That Are Good to Think
We seek, then, simple models that are good to think. These are not theories or hypotheses. Neither are they true or false in the manner of a lawful statement. Rather, they are valuable because they "generate good testable hypotheses relevant to important problems" (Levins 1966:430). A comprehensive theory about a subject is composed of a family of these models, each with its idiosyncracies. Thus, for modeling purposes "foraging" is artificially divided into "decision sets" (Krebs and Cowie 1976: 100; Pyke, Pulliam, and Charnov 1977: 140) related to such factors as diet breadth, foraging space, foraging period, and group size. Each model is uniquely relevant to a few questions and a small set of conditions. Models are not always clearly hierarchical, overlapping, or complementary. It is disconcerting to some that foraging theory works with such a motley set of individually limited tools, as if restricted applicability and the necessity to be selective were themselves a rebuke to the whole approach.


IV. Pre-State Foodways

But this confuses something like a law, which is meant to be an accurate and universal representation, with something like a structured question. If models are the latter-good to think-then better many thoughts than a few. The methodology, then, is built on (1) an optimization (or scarcity) assumption, to be distinguished from a belief that actual scarcity is a general condition; (2) a simple energy currency, tractable to those of us who can think about only a few things at a time; and (3) simple models meant to guide inquiry about a phenomenon, not to depict it.

The Diet-Breadth Model and the Marginal-Value Theorem
To exemplify this approach I will describe two key models. Consider a predator with prey scattered randomly through its environment. The diet-breadth model specifies the set (which and how many types) of resources this forager should pursue in order to achieve the highest net rate of energy intake while foraging (MacArthur and Pianka 1966; Pyke, Pulliam, and Chamov 1977: 141). The assumptions of this model are as follows:

1. The organism has the goal of maximizing its net rate of energy intake while foraging.

2. Prey are encountered randomly; the type encountered is independent of the last type encountered. 3. At a particular diet breadth, the organism always does or does not take a prey item; that is, there are no partial preferences. 4. Prey types have a stable ranking by net energy value per unit of pursuit and handling costs; that is, the quality of each prey type can be evaluated independently of other types. 5. Foraging can be divided into searching and pursuit phases, which are independent; that is, the predator searches for all prey jointly but pursues them singly. 6. Prey density is constant in a foraging interval; that is, the forager does not reach the end of its foraging range or overtake a constant rate of prey renewal during a foraging interval. 7. The benefits and costs associated with resources are treated as expected values, without relevant stochastic variance. Despite its appearance, this is not a paralyzing list. In graphic form (Figure 12.2), addition of each potential prey type reduces average search costs by an increment (LlS) that can be plotted. The average


12. The Analysis of Hunter-Gatherer Diets

FIGURE 12.2. The MacArthur and Pianka (1966) Diet-Breadth Model and Equation







Resource Rankmg NOTE: See Schoener (1974) for a definition of the terms in the equation.

pursuit and handling time per unit of prey increases by an increment (M?) because the less desirable prey added are more difficult to capture or less rewarding when secured. The optimal diet includes all ranked prey types down to the last for which the incremental savings in search costs are larger than the incremental loss due to added pursuit costs (i. e., the last before the intersection of the curves). Virtually any factors that might affect foraging can be adapted to this "systematic argument" to form a hypothesis. Thus, decreased overall habitat richness raises average search costs and expands diet breadth by stepwise addition of items of lower rank. Or, increased pursuit capabilities lower average pursuit costs, with the same effect. An especially important prediction is the following: presence of a resource in the optimal diet is independent of its abundance. It depends rather on its rank, and on the abundance of items of higher rank. The algebraic form specifies that an item x should be added to the diet only if its return relative to pursuit and handling costs (e x / tx ) is greater than the average for the diet containing all items of higher rank. In other words, the optimal forager elects to pursue an item if and only if it does not expect, in the same interval of time, to both locate and capture a more valuable item. The second model is the marginal-value theorem (Figure 12.3; Charnov 1976; Charnov, Orians, and Hyatt 1976). This model answers the question,


IV. Pre-State Foodways

FIGURE 12.3. The Marginal-Value Theorem

Net Resource Harvest

Average intake

Time in Patch NOTE: See Charnov 1976.

when should a forager leave one discrete patch to travel to another? The assumptions are:

1. The forager attempts to maximize its net rate of energy intake.

2. Food is encountered within habitat patches. 3. Patches are encountered randomly. 4. Patches are not revisited before their resources recover from exploitation. 5. Travel between one or more patch types is non-productive or lost time. 6. As an organism forages within a patch, it depresses the rate of food intake because the prey become scarce or wary, or because they flee to other locations. 7. The relevant variables are characterized by expected value. The model predicts that the optimal forager will leave a patch being harvested when the marginal capture rate there drops to the average rate of intake for the habitat. At this margin, the predator will do as well or better by seeking out an unexploited locale, even though some effort will be lost in traveling there. The


12. The Analysis of Hunter-Gatherer Diets

optimal forager abandons particular spots before prey are fully depleted, and the higher the overall quality of the environment, the sooner it abandons them.

Anthropological Studies
I have tried to convey the assumptions and constraints that give optimal foraging theory its strength and also rather strongly restrict the range of issues that it may help to resolve. Whatever its attractiveness as an exercise in structured inquiry, or problem stalking, the success of optimal foraging theory depends on the guidance it offers in actual analyses. I will restrict a brief discussion to several ethnographic analyses that use simple currency models, and I will stress relevance to important problems rather than testability of hypotheses (see also Chapter 13). Smith (1983) presents a detailed review of the ethnographic research based on foraging theory; Keene (1979, 1981) and Reidhead (1979, 1980) focus on linear programming models and archaeological research. Yesner (Chapter 11) and Cohen (Chapter 10) give additional examples of the use of the models.

Cree Foraging
The first anthropological fieldwork guided by these models was my work with the boreal forest Cree-Ojibwa in 1975 (Winterhalder 1981b, 1983a, 1983b). For a year I traveled with Cree-Ojibwa foragers, timing and assessing the energy expenditure of foraging activities, mapping on airphotos their movement through the vegetation mosaic, weighing game captured, and discussing the tactics of foraging with my companions. The results were a modest confirmation of the diet-breadth and marginal-value theorem hypotheses and a more certain demonstration that the assumptions of another model, the patch-choice model (MacArthur and Pianka 1966), were frequently violated in the CreeOjibwa case. For instance, historical changes in diet breadth give comparative evidence that fits the predictions of the diet-breadth model (Winterhalder 1981b, 1983b), and the qualitative dynamics of Cree-Ojbiwa foraging appear to confirm the marginal value theorem hypothesis. A forager nearly always leaves a localized patch of beaver, muskrat, or hare in search of another location before that patch is exhausted of capture opportunities. Equally interesting is the negative result. Using the MacArthur and Pianka (1966) model for patch choice, I predicted that habitat use would be quite generalized. In fact, not all patches containing sought game can be exploited. Anyone who has attempted to walk in


IV. Pre-State Foodways

thawed muskeg or to hunt a large, skittish herbivore like a moose in thick brush will understand immediately. And contrary to model assumptions, patches are not encountered randomly. The Cree-Ojibwa forager knows a wide landscape of vegetation communities intimately. Moreover, Cree-Ojibwa foragers do not always adopt the patch-to-patch pattern incorporated into the model. Because they know that some prey move in this manner, the foragers search for tracks in the interstices between patches, a practice that is more efficient and improves the chances of successful pursuit. Fortified with these limited successes, I suggested that foraging theory might also shed light on a key issue in hominid paleoecology (Winterhalder 1981c). Using the competitive exclusion principle, some paleoanthropologists have argued that there could be no more than one sympatric hominid species. In this view, the response of a partially cultural hominid to a like competitor would be to expand its use of resources and hence to exclude the less efficient of the two species. However, a derivative result of the diet-breadth and patchchoice models, the "compression hypothesis" (Schoener 1974), shows that although the diets of the two competitors might become more general, their use of habitat patches would specialize, producing the ecological and ultimately the evolutionary divergence necessary for joint survival. Cultural flexibility would only facilitate the selective processes for the optimal response of each species: niche divergence in the use of micro-habitats. I have argued as well that the Pleistocene overkill hypothesis of Martinthat human foragers newly arrived in North America around 11,000 B.P. caused massive extinctions of big game species (review in Webster 1981)-can be questioned using the marginal value theorem. The situation outlined by Martin-highly efficient hunters in a heterogeneous environment rich in game species and with a virtually unlimited frontier-are exactly those in which depletion of localized patches of prey is least likely. Such foragers would have moved quickly through the habitat, skimming the most easily obtained prey in each location before moving on. Each patch left behind would retain a breeding population.

Inuit Foraging and Group Size
In fieldwork with the Inuit on the northeast coast of Hudson Bay, Smith (1981) has used optimization approaches to analyze ecological determinants of huntergatherer group size and structure. The Inuit provide a classic but untested example of ecological adaptation-the winter aggregation of small hunting parties to engage in cooperative hunting of seals at their breathing holes. Since studies by Boas in the 1880s, anthropologists have noted that a seal visits


12. The Analysis of Hunter-Gatherer Diets

several breathing holes unpredictably. It has been argued that a successful harvest requires that hunters station themselves at as many of the holes as possible. By plotting group sizes against measurements of their energy efficiency, Smith was able to test this proposition. For seal hunting, his evidence shows a peak of individual efficiency at 3 hunters, corresponding to a winter camp of about 20 persons (1981: 62). This is much smaller than the 50 to 200 persons observed by Boas and others, suggesting the influence of factors besides this one. Overall, in 5 or 10 hunt types, Smith found a positive and statistically significant correlation between hunting group size and foraging efficiency. Where the energy optimization hypothesis failed, alternative factors like apprenticeship of children or partnerships to reduce foraging hazards apparently were operating.

Tropical Foragers: The Ache
Hawkes, Hill, and O'Connell (1982) have shown that foraging choices of neotropical Ache living in Paraguay match reasonably well those predicted by the optimal diet model. Whether plant or animal, a resource appears to enter the optimal diet set of the Ache by virtue of its pursuit and handling costs and the abundance of foods of higher rank as expected. That set will usually include some plants as well as animals. This answers the central question, "Why do hunters gather?", especially those that apparently could live well on meat alone. These conclusions are qualified somewhat by the Ache division of labor, in which males pass up encounters with highly ranked plants and females with highly ranked animals, lowering the average foraging of each. Where the energy-optimization hypothesis failed, alternative factors like apprenticeship of youths or partnerships to reduce foraging hazards apparently were operating.

Tropical Foragers: The Yanomamo, Ye'kwana, and Siona-Secoya
In a neotropical example, Hames and Vickers (1982) have examined data on the diet choice of Yanomamo, Ye'kwana, and Siona-Secoya foragers. These people forage radially (see Orians and Pearson 1978) from their gardens and villages. Hames and Vickers have made comparative observations on diet choice as prey are depleted adjacent to the village and as radial foraging effort grows. The observed practices of taking only more valuable game at greater distances from the village are broadly in accord with predictions. Resolution of disputes (Ross


IV. Pre-State Foodways

et al. 1980) about the effects of ecology on the lifeways of tropical horticulturalists may well come from interpretations and data like these.

Marine Fishermen in New Jersey
Finally, McCay's (1981) analysis of a marine inshore fishery on the east coast of the United States is the first explicitly to set optimal foraging theory into the context of a market economy. Some optimal foraging predictions were consistent with harvest decisions made by the fishermen, but these were a rather restricted part of a complex of choices influenced by personal values, economics, and politics. McCay argues that New Jersey fishermen are foragers, but they exist in an enveloping political economy that rather strongly restricts the insights available from foraging theory alone.

Opportunity-Cost Models
The above-mentioned studies are instances in which an optimal foraging approach has resulted in different, more complete, or better-substantiated conclusions about hunter-gatherer subsistence practices. Are there dangers in this type of exercise? I see several. It is easy to overextend the purview of optimal foraging models, or to misapply them, individually or collectively. The emaciated expository style of theoretical ecologists often does not make assumptions and constraints accessible to those of us who must be enlightened by the prose between the equations. Collectively the models are bound by the individual-level food acquisition focus and the restriction to a material currency. In their present form, they help to examine food procurement but do not readily extend to supra-individual aspects of economy, such as exchange, wherein social and cultural valuation of goods becomes especially important. And they do not have the character of lawful statements. For the remainder of this paper I will address two questions mentioned earlier-total foraging effort and risk-while trying to balance among these dangers. This means recognizing limits even while trying to push the models beyond their present limitations. I am interested first in the optimization statement stipulation: "while foraging."

The Opportunity-Cost Foraging Model
The diet-breadth model identifies the prey set giving the maximum net rate of energy intake while foraging. It does not indicate how long the organism should forage, nor how much energy it should harvest. Establishing the latter requires


12. The Analysis of Hunter-Gatherer Diets

that foraging be evaluated relative to other behaviors. Foraging takes time away from those activities; it also provides the energy needed to engage in them. I have been working on a model that addresses the question of how long to forage (see Winterhalder 1983c). It begins with several definitions. The opportunity cost of an activity (X) is the value of some alternative activity forgone because resources were invested inX rather than the alternative. This concept assumes that resources are limited relative to potential uses, that different courses of action are evaluated relative to one another, and that one activity among all the foregone activities can be identified as providing the relevant substitute measure. Marginal value recognizes that the value (or cost) of an activity can change as a function of its duration. An economic decision to cease or continue will focus on the value (or cost) of the next, the marginal, unit of that behavior (see Charnov 1976). Finally, an indifference curve connects points of equal utility or preference; a value isocline points of equal sociocultural valuation; and a fitness isocline points of equal fitness. Foraging is represented in the right-hand portion of this model (Figure 12.4); all non-foraging activities are aggregated in the left-hand portion. The righthand vertical axis rrleasures the net acquisition of energy (E a) while foraging; the left-hand vertical axis measures the total energy expenditure (El') in nonforaging activities. For the interval considered, acquisition must balance expenditure (E a == E e). The horizontal axis extending from the right shows the time spent foraging (Tj ); that extending from the left shows the time spent in nonforaging activities (Tnj). Note that increasing foraging times are read from right to left, and increasing non-foraging times from left to right, subject to the constraint that T( + T n ( == Tf' where T t represents an appropriate interval for analysis. . Combinations of time and energy consumption in non-foraging activities that have the same utility are connected by the indifference curves. Initially, take the configuration of the curves to be consistent with the standard conventions of micro-economic theory (e. g., Stonier and Hague 1972): • They are convex to the origin, implying that the scarcer of two jointly used resources is likely to be the more valuable. The marginal value of activity X increases (or its relative rate of substitution for Y decreases) as the quantity of Y grows. Thus, an organism with much non-foraging tirrle but relatively little expendable energy will probably value an additional unit of energy more than one of time. • The indifference (or isocline) map is independent of foraging intake. • Indifference curves or isoclines more distant from the origin have higher value. In principle, the capacity for increased utility is insatiable. • And the forager is knowledgeable, has stable preferences, and acts rationally by attempting to obtain the greatest possible utility.


IV. Pre-State Foodways

FIGURE 12.4. The Opportunity-Cost Foraging Model for a Time-Minimizer

\. *


NOTE: The asterisk (*) shows the optimal foraging time allocation (Tf ) and energy acquisition (£a) for two foraging constraint lines, A and B. The vertical bars below the X-axis show how time IS partitioned between non-foraging activities (Tnf ) and foraging (Tf ) in each case.

The O-to-a line shows the minimum survivable rate of energy expenditure; the O-to-~ line shows the maximum rate attainable in non-foraging activities. A set of indifference curves can be very sensitive to a forager's circumstances. The model offers distinctions that pass continuously from the strictly time-limited to the strictly energy-limited cases. If the marginal gain in utility from a unit of time freed from foraging is greater than that attached to an additional unit of energy gained by foraging, then the organism is time-limited and will act as a "time-minimizer" (Schoener 1971.). The indifference curves will tend toward the vertical (Figure 12.4). Conversely, if the marginal gain associated with a unit of energy is larger, the organism is energy-limited (i. e., it will act as an "energy-maximizer") and has a indifference map that tends toward the horizontal (Figure 12.5). A forager may act as an energy-maximizer in one portion of its indifference map and as a time-minimizer in another. It may switch from one to the other with seasonal or other temporal factors. The left and right portions of this model are linked in the following way. Each foraging option is associated with a net gain of energy (E a) for each unit of foraging time (Tf ). This is represented by a foraging-constraint line on the indifference map: The optimal time investment (and thus the optimal energy


12. The Analysis of Hunter-Gatherer Diets

FIGURE 12.5. The Opportunity-Cost Foraging Model for an EnergyMaximizer



'- '-

" "-






' - "'-=....,.


"~ "" " "


T ~ ~~~'i--'---~ft""""i "T"i"t-r-t-t1-t',I-iI-jIr-t-'t-It+11+'+1-+1-+1-+1-11-11_~It-I+-I+-I+-+ 11-+1-+1--f1-t--j1~~U-l-i---!-l:...J~"
I I 1'1



--.--r-T'I 1 I I 1111111111

I ......







t -'



NOTE: The asterisk (*) shows the optimal foraging time allocation (T f ) and energy acquisition (E ,J for two foraging constraint lines, A and B. The vertical bar below the X-axis shows how tIme IS partitioned between non-foraging activitIes (n and foraging (T f ) in each case.

acquisition) is determined by the point on the foraging-constraint line that is tangent to the highest indifference curve (or value/fitness isocline). This linkage establishes the opportunity costs (time) and benefits (energy) of foraging by evaluating them relative to alternative uses in other activities. If foraging conditions vary, so will the diet choice and efficiency of the optimal hunter-gatherer. This change will alter the slope of the foraging-constraint line. Rotating that line over the indifference map (Figure 12.6) generates a consumption curve (cc) that characterizes the forager's effort and harvest for all ecological circumstances.

An Opportunity-Cost Currency
The opportunity-cost approach directs attention beyond the diet-breadth model to some related and here exploratory lines of inquiry. What is the relevant currency for the opportunity costs experienced by a human forager? Is there an


IV. Pre-State Foodways

FIGURE 12.6. An Opportunity-Cost Model that Depicts the Forager's Consumption Curve (cc)

expected form for a hunter-gatherer opportunity-cost map? What determines that configuration? Some anthropologists using foraging models (Hawkes et al. n. d.) state flatly that fitness is the sole appropriate currency. I do not find that position compelling. First, fitness has a special, limited, theoretical meaning: differential reproductive success of phenotypic traits expressed variably among the individuals of a population, for those phenotypes tied more or less directly to a corresponding inter-individual genotypic variance. Given our ignorance about mechanisms of biocultural evolution, appeals to fitness outside this biological setting are analogies carrying heavy loads of ambiguity. Second, it is nearly impossible to make a quantitative assessment of the incremental contribution of a particular phenotype/genotype linkage to relative fitness. For those seeking a rigorous, operational measure of segments of human behavior, fitness is a poor candidate. Fitness in the abstract, even the fitness of individuals, will not help, because the analysis actually considers fitnesses associated with a restricted set of activities extracted from the totality of behavior. Finally, Mayr (1974) argues that resource-acquisition behavior has evolved toward "open" behavioral programs based on generalized abilities to accumulate and interpret experience, to learn, and to make choices. It is based in the "cultural retention of


12. The Analysis of Hunter-Gatherer Diets

individually adaptive behavior" and "cognitive evaluation and retention of beneficial customs" (Pulliam 1981: 62). We should expect, then, that forating decisions are heavily influenced by sociocultural information that may be independent of fitness considerations. In a context in which it cannot be defined or measured, and in which even its relevance is at issue, invocations of fitness must entail some other kind of adaptive judgment, usually related to material utility. The alternative, however, is not particularly attractive. Biologists discomforted by ambiguities in a central concept (e. g., Stearns 1976: 4) will find company with the economists and anthropologists when it comes to preference and utility. Despite this, I propose that the appropriate currency for this opportunity-cost map is utility. In economics utility theory has a convoluted history (Page 1968). Bentham, in his own words, "planted the tree of utility . . . deep, and spread it wide" (quoted in Stigler 1965: 66) in the late 18th century, a calculus for an introspective but measurable quantity that could be used for designing efficacious moral legislation. Over the next hundred years the concept changed profoundly. Bernoulli and others formalized the idea of marginal utility. Edgeworth introduced the convention of indifference curves, here adapted to ecological analyses. Marshall got the curves firmly linked to a theory of consumer demand. Pareto showed that utility was not measurable-that, in effect, utility maps represented only ordinal relationships of preference. Hicks and Allen argued that introspective ideas like utility were unnecessary if one replaced marginal utility with marginal rate of substitution. And Samuelson completed the program by showing that behaviorist analysis of "revealed preference" freed the theory of all unscientific notions, like utility. The utility in utility theory extinguished itself. From a social theory it has been transformed into an idea strictly tied to price-demand theory, ordinal, thoroughly behaviorist, and therefore presumably free of any base sociocultural content or psychological assumptions. Despite the Nobel Prize attached to the end of this sequence, I am most comfortable with the utility of the pre-Samuelson period (see Sen 1977; Wong 1978). I define utility as the worth people attribute to things or actions, either because they are useful, in an adaptive or fitness-enhancing sense, or because they have status within sociocultural systems of meaning or exchange. Preference is the expression of such rankings, revealed in choice or evident in action. Fitness draws on the deductive power of evolutionary theory; sociocultural valuation on the surety that learned beliefs and experience exert a strong and more or less independent influence on the variability in the specific forms of human foraging choices. Both are necessary for human adaptation research, but neither in the abstract is very helpful. To interpret them (as isocline maps) or evaluate their combined action (in the form of utility), we are forced to consider empirical understandings of behavior and environment, to generalize from facts about the meaning of our deductive postulates, to inquire about,


IV. Pre-State Foodways

observe, and, if need be, assume preferences. This is necessary to translate those abstract currencies into an indifference map that will produce interesting, testable hypotheses. Utility and preference are messy concepts, but until we have sound theories of biocultural evolution, their use is unavoidable.

An Idealized Hunter-Gatherer Indifference Map
I will also propose a nonnative or idealized form for the indifference map of a hunter-gatherer. Consider the following question: How would we represent the case in which there is a social stigma or material disadvantage associated with too much or too little success in the harvesting of food? Note that this allows social phenomena an explicit influence on individual preference. Note also that this map will portray long-tenn averages or behavioral tendencies, ignoring day-to-day vagaries of motivation and success in the food quest. Social controls come into effect when a person exceeds or falls short of a consumption nonn. Controls might have three effects. First, they might cause the person to reevaluate his or her preferences toward greater confonnity with the nonn. Second, if the person adamantly holds to certain preferences, the group, by withdrawing assistance or otherwise impeding the perfonnance of non-foraging tasks, could reduce that individual's ability to derive utility from various time and energy combinations. Finally, they might reduce the foraging efficiency of the recalcitrant, perhaps by withholding information, thus depressing his or her foraging-constraint line and reducing energy intake (if not the commitment to consumption). Graphically, a social consumption nonn will take the fonn of a line intermediate between minimum and maximum non-foraging expenditure rates. It can be derived as follows. Begin with a garden-variety indifference map (Figure 12.7), with an individual consumption curve different from the social consumption nonn. Social pressures (the second alternative above) can be represented by arrows transfonning the indifference map. Above the consumption norm these cause the indifference curves to pivot upward and to the right because the utility that the person can derive from various time-energy combinations is lowered. Below the consumption norm, disapproval of insufficient harvest will also lower the utility of various time-energy combinations, again pivoting the indifference curves upward and to the right. Whether this individual begins as a time-minimizer or an energy-maximizer, his or her individual consumption curve will approach the accepted nonn. The indifference map generated by this exercise is an intuitively reasonable hypothesis about individual preferences as such. If energy shortfalls approach sustenance levels, tactics will tend toward energy maximization and longer foraging. If intake grows to or above a comfortable level, people will tend to


12. The Analysis of Hunter-Gatherer Diets

FIGURE 12.7. Generation of a Consumption Norm (cn) by Group Pressure on an Individual's Preferences (Indifference Map)











~ ~,



adopt time-minimizing strategies and engage in less foraging. This hypothesis also recognizes that the coercive powers of fitness and sociocultural factors over human behavior are not evenly distributed over the indifference map. In the lower and left-hand portions, for instance, sufficient food assumes more importance. Health and survival (and perhaps, but not necessarily, fitness) may be at stake. In the upper and right-hand portion, basic needs have been satisfied, and historically specific, sociocultural determinants of needs and wants may prevail. People are guided sometimes by the belly and sometimes, quite independently, by the head. This idealized indifference map should be viewed as a baseline hypothesis. The actual map of any particular forager or foraging group may be different in ways that must be given a situational explanation. Deviations will be a function of various factors, including the following: 1. The number and value of the non-foraging activities performed, including the support of dependents, and their collective demands on energy and time 2. The potential for satiety with respect to energy and time, and their use in non-foraging activities 3. Differences in age, sex, and individual competence in non-foraging tasks


IV. Pre-State Foodways

4. Physical hazards and perceived comforts or discomforts of foraging, relative to those for non-foraging activities 5. Uses of resources aside from immediate consumption by foragers or their dependents (perhaps in exchange or for storage) 6. Temporal and situational factors that influence the opportunity costs of behaviors through scheduling constraints on alternative activities This last point is quite important. The "cost" of an activity is not necessarily the actual time or energy expended on it. For instance, tool manufacture in the secure warmth of one's hut, after the evening's reproductive fitness has been secured and while hunting and gathering are on hold until the following day, has no cost except the increment of energy expenditure above the resting metabolic rate.

Economists have no qualms about the assumption that human wants for goods are unlimited, that ever greater amounts of utility can be gained from ever larger baskets containing, for instance, potatoes and grapes (Stonier and Hague 1973: 53). The notion of surfeit is absent from the workings of microeconomics, which has formalized the ignorance of excess. But even if we accept insatiability for combinations of goods in a market economy, the combination of time and energy represented in a hunter-gatherer indifference map is a different matter. The time of the model refers to a finite interval, and the metabolic use of energy has a limit at the point of exhaustion. The foraging mode may itself place structural constraints on the use of energy (Lee 1981). Thus, for foragers, consumption of energy is tied to physiological and socioeconomic constraints quite unlike those affecting the consumption of goods in a market economy (where the object is not acquisition of potatoes and grapes to eat but competitive transformation of commodities to sustain a profit or retain a wage). This brings us to Sahlins's (1972) proposition that hunter-gatherers are in business for their health (a not un-Darwinian statement), and a new question: How would satiation affect the model? It is possible and perhaps likely that foragers have, and readily attain, an upper bound for energy/time utility, and express it in their preferences. If non-foraging leads to boredom, which impels the forager to hunt or gather, or if the social constraints on material acquisition or aggrandizement through accomplishments cause an unusually successful forager to forgo the food quest temporarily (R. B. Lee, pers. comm.), then maximum utility may actually be a curved ridge, or it may reach a plateau (Figure 12.8). More time foraging and greater energy harvest may reach a point of


12. The Analysis of Hunter-Gatherer Diets

FIGURE 12.8. An Indifference Map with a Utility/Fitness Plateau, Generating a Consumption Space (cs)

disutility or an indeterminant zone of like utility. A ridge would cause the consumption curve to bifurcate into two optimal harvest options; a plateau would generate a set of equivalent options, a consumption space. There is no longer a single equilibrium, but a set of alternatives among which pressures for adaptive optimization carry no determinancy. These latter two indifference maps (Figures 12. 7 and 12.8) are important because they provide a link between optimal foraging theory and some more traditional analyses of foragers, including the "original affluence" and "limited needs" proposals of Sahlins (1972) and Lee (1979; cf. Hawkes and O'Connell 1981). It is a revealing link because some (Hawkes et a1. 1985) have claimed, incorrectly I believe, that an evolutionary ecology approach is theoretically incompatible with this earlier work. Sahlins and Lee inquire principally about the political economy of foraging, raising questions that cannot be answered by foraging theory per see "Original affluence" and "limited needs" are mainly concepts of sociocultural valuation. They enter an opportunity-cost foraging analysis through their effect on the shape of the individual's indifference map. To have limited needs with respect to foraging is to be, in effect, a time-minimizer. A forager may seek to be optimal in the choice of a foraging strategy in order to be a time-minimizer and energy satisfier in the food quest. There are no theoretical impediments to this possibility, nor does it seem unlikely, given


IV. Pre-State Foodways

the ethnographic record of hunter-gatherer preferences and foraging efficiencies. The proposition that there is always more utility to be gained from hard productive work is more aptly ascribed to Calvin than to Darwin.

Risk-Sensitive Foraging Models
A second body of recent work focuses on the effects of stochasticity in foraging decision variables (e. g., encounter rate). Most existing models, including those discussed above, are deterministic, based on the simplifying assumption that expected (or average) values adequately characterize the important factors in foraging decisions. But, in fact, foragers may respond to the mean and variance in the rewards associated with each diet choice. They may be "risk-sensitive," depending on their preference for variance or certainty in combination with differing mean food reward expectations. A simple result of this work is the prediction that a forager will make choices that avoid variance if its current energy budget is positive, but seek variance if that budget is negative. Figure 12.9 is a simplified depiction of this prediction. There are four possible foraging decisions (A through D), each with a different mean/variance combination in the associated food reward. For positive mean expected intakes (u t ), those above the net minimum survivable requirement (R net), minimizing variance (vart ) is optimal (A is preferable to B); for u t below R ner maximizing variance will result in at least some positive outcomes (D is preferable to C). Stephens and Chamov (1982) have derived a model that generalizes the results depicted in Figure 12.9. The probability of starvation (Z') may be given as follows (in a form that avoids mathematical details to the extent possible):

Risk diminishes as Z' decreases, with the troublesome mathematical quirk that the organism is in positive energy balance when Z' is negative. Reference to the equation will show that for positive values of Z'
R net

ut > 0

That is, there is a negative energy balance; increasing variance decreases risk. For negative values of Z'

R net


ut < 0


12. The Analysis of Hunter-Gatherer Diets

FIGURE 12.9. The Extreme Variance Rule for Risk-Sensitive Foraging Decisions in a Stochastic Environment

+ Energy Balance
- - - R net

- Energy Balance





NOTE: Each of the foraging choices (A through D) has associated with it a distribution of food rewards, characterized by a mean and a variance.

There is a positive energy balance; decreasing variance decreases risk. Large expected values relative to minimum requirements (R net - u t ~ 0; Z' is a large negative number) also reduce risk. The equation can be reorganized as:

the standard slope-intercept equation, and then depicted graphically as in Figure 12.10. The full set of mean/variance combinations associated with the available foraging choices can be represented by points (varp u t ) on this plane. The optimal diet is that combination intersected by the line extending from R net with the greatest slope. (This is because the greater the slope (-Z'), the lower the value of Z, which we want to minimize.) Stephens and Charnov derive the mean/variance combinations for the marginal-value theorem that result from stochastic variance in the time required to locate a patch. They get an oblong curve with these properties: (1) the


IV. Pre-State Foodways

FIGURE 12.10. Generalized Stephens/Chamov (1982) Model for Foraging Decisions with Stochastic Elements


1\ 1\


/ y t

Mean Energy Gain





R n ('!








Standard Dematwn of Energy Gain
NOTE: The greatest slope (- Z') of the line that intersects one of the available mean/vanance paIrs (shaded area) mimmizes Z' and hence risk.

stochastic result will usually approximate the deterministic outcome; but (2) if minimum requirements are less than the expected maximal intake predicted by the deterministic model, then the forager should remain in patches somewhat longer than predicted by the deterministic model. That the complicated stochastic outcome approximates the simple deterministic one is encouraging, but the unusual set of mean/variance combinations that gives rise to this consequence suggests caution in extrapolating to other situations.

Conclusions on the Process of Stalking
To summarize, optimal foraging models may provide one answer to Bettinger's (1980) "impending crisis" in hunter-gatherer studies-the need for new "predictive/explanatory" models to guide research and interpretation. They may


12. The Analysis of Hunter-Gatherer Diets

provide a "consequence law"-in the restricted sense of Cohen (1978)-and hence a method for mitigating some shortcomings of functionalism (see Smith and Winterhalder 1981). By attention to the history of the logistic equation and by carefully listing the assumptions of optimal foraging theory, I hope to have conveyed a particular and somewhat cautious attitude toward models. Partial skepticism is the best favor one can do a favorite theory. The fieldwork examples were meant to demonstrate that this is a heuristically rewarding type of inquiry. The section on opportunity costs and risk was meant to further develop the theory, and to define its boundaries. Extensions of opportunity-cost models may allow us to use formalist approaches in hunter-gatherer foraging and diet studies even while clinging to substantivist insights on forager goals. The initial results of risk-sensitive analyses suggest that we may be able to retain the simplicity of some deterministic models even while acknowledging the presence of stochasticity. In combination with fieldwork, these and other developments using evolutionary ecology theory may eventually generate a rigorous understanding of the ecological component of the foraging mode of production.

References Cited
Bettinger, R. L. 1980 Explanatory IPredictive Models of Hunter-Gatherer Adaptation. Advances in Archaeological Method and Theory 3: 189-255. Charnov, E. L. 1976 Optimal foraging: T'he Marginal Value 1'heorem. Theoretical Population Biology 9: 129-36. Charnov, E. L.; G. Orians; and K. Hyatt 1976 Ecological Implications of Resource Depression. Anlerican Naturalist 110:247-59. Cody, M. L. 1974 Optimization in Ecology. Science 183: 1156-64. Cohen, G. A. 1978 Karl Marx's Theory of History: A Defence. Princeton, N.].: Princeton University Press. Connell,]' H. 1975 Some Mechanisms Producing Structure in Natural Communities: A Model and Evidence from Field Experiments. In Ecology and the E volution of Communities, M. L. Cody and]. M. Diamond, eds., pp. 460-90. Cambridge: Harvard University Press. Durham, W. H. 1976 The Adaptive Significance of Cultural Behavior. Human Ecology 4:89121.


IV. Pre-State Foodways

Gause, G. F. 1934 The Struggle for Existence. Baltimore: Williams and Wilkins. Gould, S. ]., and R. C. Lewontin 1979 The Spandrels of San Marco and the Panglossian Paradigm: A Critique of the Adaptationist Programme. Proceedings of the Royal Society ofLondon B 205:581-98. Hames, R. B., and W. T. Vickers 1982 Optimal Diet Breadth Theory as a Model to Explain Variability in Amazonian Hunting. American Ethnologist 9: 358-78. Hawkes, K.; K. Hill; and J. F. O'Connell 1982 Why Hunters Gather: Optimal Foraging and the Ache of Eastern Paraguay. American Ethnologist 9:379-98. Hawkes, K., and]. F. O'Connell 1981 Affluent Hunters? Some Comments in Light of the Alyawara Case. American Anthropologist 83: 622-26. Hawkes, K.; J. F. O'Connell; K. Hill; and E. Charnov 1985 How Much Is Enough: Hunters and "Limited Needs. Ethology and Sociobiology 6: 3-15. n. d. Why Reduce It to Fitness? Optimal Foraging and Hunter-Gatherers. Manuscript. Hutchinson, G. E. An Introduction to Population Ecology. New Haven: Yale University 1978 Press. Jacob, F. 1977 Evolution and l'inkering. Science 196: 1161-66. Jochim, M. 1982 Optimization and Risk. Manuscript. Johnson, A., and C. A. Behrens 1982 Nutritional Criteria in Machiguenga Food Production Decisions: A Linear-Programming Analysis. Human Ecology 10: 167-89. Keene, A. S. 1979 Economic Optimization Models and the Study of Hunter-Gatherer Subsistence Settlement Systems. In Transformations: Mathematical Approaches to Culture Change, C. Renfrew and K. Cooke, eds., pp. 369404. New York: Academic Press. 1981 Optimal Foraging in a Nonmarginal Environment: A Model of Prehistoric Subsistence Strategies in Michigan. In Hunter-Gatherer Foraging Strategies: Ethnographic and Archaeological Analyses, B. Winterhalder and E. A. Smith, eds., pp. 171-93. Chicago: University of Chicago Press. Kingsland, S. 1982 The Refractory Model: The Logistic Curve and the History of Population Ecology. Quarterly Review of Biology 57: 29-52. Krebs, ]. R., and R. ]. Cowie 1976 Foraging Strategies in Birds. Ardea 64:98-116.


12. The Analysis of Hunter-Gatherer Diets

Lee, R. B. 1979

Levins, R. 1966

The !Kung San: Men, Women and Work in a Foraging Society. Cambridge: Cambridge University Press. Is There a Foraging Mode of Production? Canadian Journal of Anthropology 2: 13-19.

The Strategy of Model Building in Population Biology. American Scientist 54:421-31. Lewontin, R. C. 1978 Fitness, Survival and Optimality. In Analysis ofEcological Systems, D. j. Hom, G. R. Stairs, and R. D. Mitchell, eds., pp. 3-21. Columbus: Ohio State University Press. MacArthur, R. H., and E. R. Pianka 1966 On Optimal Use of a Patchy Environment. American Naturalist 100:603-9. McCay, B. 1981 Optimal Foragers or Political Actors? Ecological Analysis of a New Jersey Fishery. American Ethnologist 8:356-82. Maynard Smith, j. 1978 Optimization Theory in Evolution. Annual Review of Ecology and Systematics 9:31-56. Mayr, E. 1974 Behavior Programs and Evolutionary Strategies. American Scientist 62:650-59. Orians, G. H., and N. E. Pearson 1978 On the Theory of Central Place Foraging. In Analysis ofEcological Systems, D. j. Hom, G. R. Stairs, and R. D. Mitchell, eds., pp. 155-77. Columbus: Ohio State University Press. Orlove, B. S. 1980 Ecological Anthropology. Annual Review ofAnthropology 9:235-73. Page, A. N. 1968 Utility Theory: A Book ofReadings. New York: John Wiley. Pulliam, H. R. 1981 On Predicting Human Diets. Journal ofEthnobiology 1:61-68. Pyke, G. H.; H. R. Pulliam; and E. L. Chamov 1977 Optimal Foraging: A Selective Review of Theory and Tests. Quarterly Review ofBiology 52: 137-54. Rapport, D. j. 1980 Optimal Foraging for Complementary Resources. American Naturalist 116:324-46. Reidhead, V. A. 1979 Linear Programming Models in Archaeology. Annual Review of Anthropology 8: 543- 78. 1980 The Economics of Subsistence Change: Test of an Optimization Model.


IV. Pre-State Foodways

In Modeling Change in Prehistoric Subsistence Economies, T. K. Earle and A. L. Christenson, eds., pp. 141-86. New York: Academic Press. Richerson, P. ]. 1977 Ecology and Human Ecology: A Comparison of Theories in the Biological and Social Sciences. American Ethnologist 4: 1-26. Ross, E. B.; ]. B. Ross; N. A. Chagnon; and R. B. Hames 1980 Amazonian Warfare (Letter). Science 207:590-93. Sahlins, M. 1972 Stone Age Econornics. Chicago: Aldine. Schoener, T. W. 1971 Theory of Feeding Strategies. Annual Review of Ecology and Systematics 2: 369-404. The Compression Hypothesis and Temporal Resource Partitioning. Pro1974 ceedings of the National Academy ofSciences, USA 71: 4169-72. 1982 The Controversy Over interspecific Competition. American Scientist 70:586-95. Sen, A. K. 1977 Rational Fools: A Critique of the Behavioral Foundations of Economic Theory. Philosophy and Public Affairs 6:317-44. Smith, E. A. 1979 Human Adaptation and Energetic Efficiency. Human Ecology 7: 53- 74. 1981 The Application of Optimal Foraging l'heory to the Analysis of HunterGatherer Group Size. In Hunter-Gatherer Foraging Strategies, B. Winterhalder and E. A. Smith, eds., pp. 36-65. Chicago: University of Chicago Press. Optimal Foraging T'heory and Hunter-Gatherer Societies. Current An1983 thropology 24: 625-51. Smith, E. A., and B. Winterhalder 1981 New Perspectives on Hunter-Gatherer Socio-Ecology. In Hunter-

Gatherer Foraging Strategies: Ethnographic andA rchaeological Analyses, B. Winterhalder and E. A. Smith, eds., pp. 1-12. Chicago: University of
Chicago Press. Speth, ]. D., and K. A. Spielmann 1983 Energy Source, Protein Metabolism, and Hunter-Gatherer Subsistence Strategies. Journal ofAnthropological Archaeology 2: 1-31. Stearns, S. C. 1976 Life History Tactics: A Review of the Ideas. Quarterly Review ofBiology 51:3-47. Stephens, D. W., and E. L. Charnov 1982 Optimal Foraging: Some Simple Stochastic Models. Behavioral Ecology and Sociobiology 10:251-63. Stigler, G. ]. The Development of Utility Theory. In Essays in the History ofEconom1965 ics, G. ]. Stigler, ed., pp. 66-155. Chicago: University of Chicago Press.


12. The Analysis of Hunter-Gatherer Diets

Stonier, A. W., and D. C. Hague 1973 A Textbook of Economic Theory. Ne\\T York: Wiley. Webster, D. 1981 Late Pleistocene Extinction and Human Predation: A Critical Overview. In Omnivorous Primates: Gathering and Hunting in Human Evolution, R. S. O. Harding and G. Teleki, eds., pp. 556-94. New York: Columbia University Press. Westoby, M. 1978 What Are the Biological Bases of Varied Diets? American Naturalist 112:627-31. Wiens, J 1977 On Competition and Variable Environments. American Scientist 65: 59097. Winterhalder, B. 1981a Optimal Foraging Strategies and Hunter-Gatherer Research in Anthropology: rrheory and Models. In Hunter-Gatherer Foraging Strategies: Ethnographic and Archeological Analyses, B. Winterhalder and E. A. Smith, eds., pp. 13-35. Chicago: University of Chicago Press. 1981b Foraging Strategies in the Boreal Forest: An Analysis of Cree Hunting and Gathering. In Hunter-Gatherer Foraging Strategies, B. Winterhalder and E. A. Smith, eds., pp. 66-98. Chicago: University of Chicago Press. 1981c Competitive Exclusion and Hominid Paleoecology: Limits to Similarity, Niche Differentiation and the Effects of Cultural Behavior. Yearbook of Physical Anthropology 24: 101-21. 1983a History and Ecology of the Boreal Zone in Ontario. In Boreal Forest Adaptation: The Algonkians of Northern Ontario, A. 1~ Steegmann, Jr., ed., pp. 9-54. New York: Plenun1 Press. 1983b Boreal Foraging Strategies. In Boreal Forest Adaptations: The AIgonkians of Northern Ontario, A. T. Steegmann, Jr., ed., pp. 201-41. New York: Plenum Press. 1983c Opportunity Cost Foraging Models for Stationary and Mobile Predators. A merican Naturalist 122: 73-84. Wong, S. 1978 The Foundations of Paul Samuelson's Revealed Preference Theory. London: Routledge and Kegan Paul.



How Much Food Do Foragers Need?
i\N INTERESTING EMPIRICAL PUZZLE IS POSED BY RECENT OBSERvations on foraging effort among hunter-gatherers. In the mid-1960s, the Dobe ~Kung of Botswana reportedly spent no more than 20 hours per week collecting resources that yielded an average of 2,100 kilocalories per consumer day (Lee 1968, 1969). Many anthropologists take these figures to be typical of hunter-gatherers. They expect that work effort is set to meet some general and specifiable consumption need, so that when this need is easily met, foragers work little. When it takes more effort to reach the same goal, workloads increase. Yet, in contrast to Lee's reports on the !Kung, the Ache of eastern Paraguay have recently been observed to work, averaging male and female foragers, at least 35 hours per week for a mean return of 3,800 kilocalories per consumer day (Hill et aI., 1984). They procure more food than Lee reports for the! Kung, and they work longer hours to get it. The Ache data are inconsistent with what has become the received wisdom on hunter-gatherer foraging, and they raise an obvious and important theoretical question: what determines the amount of time hunter-gatherers devote to foraging and the quantity of food they consume?

The Problem
Largely because of the rich ethnographic descriptions (e. g., Lee 1968, 1979; Marshall 1976; Thomas 1959; and films by John Marshall), the !Kung are often identified as prototypical hunter-gatherers-well-nourished foragers who work relatively short hours, rely on the widely available plant foods collected by women for the bulk of their diet, and maintain low birthrates, which prevent population growth from threatening local resources. Though this characterization has been challenged, as applied both to the! Kung themselves and to other hunter-gatherers (e.g., Ember 1978; Lee and DeVore 1968; Truswell and


IV. Pre-State Foadways

Hansen 1979; Wilmsen 1978), it remains remarkably persistent. When anthropologists write about hunter-gatherers, more often than not they write about the !Kung (e.g., Cohen 1977; Harris 1977, 1983; Leakey and Lewin 1978). The Ache are very different. The per capita daily consumption rates they sustain while foraging are quite high, averaging 2,600 to 5,400 kilocalories, depending on the season (Hill et al. 1984). More than half this total comes in the form of meat. This is especially surprising in view of generalizations about tropical hunter-gatherers (Lee 1968), particularly those of the South American lowlands (Lathrap 1968). These figures are not the result of unusually high local capture rates. Hourly returns for Ache hunters average only 10 to 15 percent higher than those calculated for the !Kung (Hawkes, Hill, and O'Connell 1982; Hill and Hawkes 1983). The difference in return is almost entirely a function of the time Ache men devote to hunting-nearly 50 hours per week per hunter while on foraging trips, more than twice the figure reported by Lee (1969) for !Kung men. The contrast becomes even more striking when one considers the difference in women's work effort. !Kung women spend about 12 to 19 hours per week collecting food (Lee 1968; but see recalculation by Hawkes and O'Connell 1981); yet for Ache women, the figure seldom exceeds 10 hours per week. Stimulated by Lee's description of the !Kung, Sahlins characterized huntergatherers as "the original affluent society" (1968, 1972). Wanting little, they could work little to have all their needs met. Yet neither the differences in overall time investment in foraging between the !Kung and the Ache nor the differences in relative time investment between men and women show the pattern that might be expected if some fixed food total were the goal. The !Kung and Ache are both of small stature. They are of about the same height. Vet the Ache eat a great deal more, and they are almost 20 percent heavier (Hill et al. 1984; Howell 1979). If needs were a function of some standard nutritional goal, patterns of caloric intake should be more similar than they apparently are. Moreover, the Ache are not an isolated case, an oddity among huntergatherers with respect to foraging. Other groups described within the past few years (Harako 1981; Hawkes and O'Connell 1981; Tanaka 1980), including the !Kung themselves (Yellen 1977), display patterns of work effort and food consumption quite different from those reported by Lee. These data underline the problem of interest here: how much food will hunter-gatherers eat, and how much time will they spend getting it? Recommended daily allowances and minimum daily requirements do not provide answers to these questions. To label the differences as "cultural" only begs the question. Can we account for the variation in some nomothetic way that will allow us to predict other cases, including patterns in the past?


13. How Much Food Do Foragers Need?

A Sociobiological Perspective
There is wide agreement among anthropologists that vanatIon in huntergatherer behavior is, in some sense, an adjustment to local circumstancesthat it is "adaptive." However, opinions differ about the meaning of adaptation. If the direction and the means of adjustment are not specified, adaptation is an empty notion. Supposing that people who depend on local environments suit their pattern of exploitation to the resources available gives us no way to predict which plants and animals they will use as resources, how much they will crop, or what patterns of movement they will display. As circumstances change, what sort of adjustment to suit these changes do we expect? Without some specification of the goal of adjustments, the direction and form they will take cannot be predicted. However, a strict biological notion of adaptation does specify a goal: maximizing fitness. Adaptations are features of living organisms spread by natural selection (Williams 1966), that is, features spread according to the fitness-the relative reproductive success-they confer. If we can take advantage of this rigorous notion of adaptation, we have a powerful explanatory and predictive theory with which to work. The warranting argument is straightforward and simple. Past selection has shaped capacities and tendencies for enormous flexibility in humans. This flexibility spread in ancestral populations because it was adaptive; it provided a fitness advantage to its carriers. This means it is a flexibility of a particular kind-not flexibility to do anything at all, but flexibility to track locally changing fitness opportunities. Tendencies to do otherwise would lower reproductive success and so continually be extinguished whenever they appeared. If this is so, the changes, the differences, we see in human behavior over time and space need not be the consequence of changing gene frequencies to allow explanation in terms of natural selection. They are the consequence of individuals adjusting to take advantage of changing fitness options. This means that we can use a strict biological definition of adaptation to apply to patterns of human behavior and a fortiori the behavior of hunter-gatherers. Adaptations are features that maximize the (inclusive) fitness of those displaying them. This allows us to set questions about variation in the behavior of hunter-gatherers in a fundamental theory that applies not only to Homo sapiens and not only to behavior but to all the changes in hominids throughout the Pleistocene. From this perspective the key question in the analysis of any particular behavioral trait is how that trait contributes to the inclusive fitness of the individuals who display it. To address this question, one must construct a series of specific hypotheses about the fitness-related goal of the behavior in question, the alternative means available to achieve that goal, and the costs and benefits associated with each.


IV. Pre-State Foodways

It is important to note that in such an analysis neither adaptation, selection, nor optimality is directly under test (Maynard Smith 1978). They provide the theoratical motivation for and links of coherence among various hypotheses. Nor is fitness itself usually measured directly. Rather, attention is directed to "design for fitness" (Williams 1966)-that is, to measurable criteria, such as energetic efficiency, that are likely to be closely related to fitness (Pyke, Pulliam, and Charnov 1977; Smith 1979). If we apply this perspective to the analysis of food acquisition patterns among the !Kung, Ache, and any other hunter-gatherers, our attention is immediately directed to the fitness-related costs and benefits of foraging-that is, to the relative advantages and disadvantages of alternative foraging strategies under different circumstances. To the extent that we can correlate these fitness costs and benefits to features of local ecology, we develop understanding applicable not only to the present but to the Pleistocene as well.

The !Kung
It is ironic that Lee's work on the! Kung, which set a new standard for ecological anthropology, seems to show that !Kung behavior is inconsistent with a biological definition of adaptation. !Kung women, so the data suggest, work relatively little, never exhaust local resources, yet have comparatively few children. This is paradoxical. If there is plenty of food available and plenty of time to acquire it, they could comfortably feed more children than they have. Under these circumstances any tendency to shorten the birth interval and increase the number of offspring would increase parents' fitness. Why don't the !Kung work harder, eat more, and turn more resources into babies? Could it be that in spite of appearances the !Kung are working as hard and eating as much as will give them the maximum number of grandchildren? Following a lead from Lee (1972), Blurton Jones and Sibly (1978) asked precisely this question some time ago in an essay that remains curiously underappreciated. They measured the labor costs of alternative inter-birth intervals given five assumptions: (1) that women seasonally provide about 60 percent of family caloric intake in the form of mongongo nuts; (2) that mothers usually take children under the age of four on foraging trips; (3) that these children depend heavily on mother's milk; (4) that women could not gather nuts daily because of the processing time these nuts require; and (5) that dry-season camps must be up to six miles from nut groves because of the distribution of groves and permanent water holes. Blurton Jones and Sibly found that with these constraints, the maximum backload that did not threaten thermal ex-


13. How Much Food Do Foragers Need?

haustion is sustained by carrying the nut-and-baby load set by a four-year birth interval. In other words, !Kung women seem to be working as hard as they safely can in order to get food to support as many children as possible under the prevailing ecological circumstances. The apparent challenge to a biological perspective is provisionally rebutted, at least as far as women's work effort and reproductive strategy are concerned. But what about men's work effort? If more food means more surviving offspring, why don't !Kung men hunt more frequently? It would certainly seem as if their wives should try to persuade them to do so. Can it be that the low level of hunting effort is one that maximizes men's reproductive success given the options available? If Blurton Jones and Sibly are right, the heat and humidity constraints that limit women's work effort may impede the men's as well. Although hunters do not carry heavy burdens, except when bringing game to camp, the faster pace they sustain in tracking and pursuit may increase the danger of heat exhaustion and raise water requirements. Lee (1979: 105-7) notes that "a person walking in the sun at 38°C (100°F) sweats at the rate of roughly 800 cc of water loss per hour, an equivalent of over 3 liters (over 6lb of water) in a typical working day." If either high temperatures or a relative lack of water limit men's hunting efforts, we might expect to see a pattern of seasonal variability in this activity. Lee (1968, 1969, 1979) presents quantitative data on hunting for a four-week period in July-August, the coolest and nearly the driest time of year. He says that tracking conditions during this period "encourage more hunting and the setting up of snare lines" (Lee 1979: 104), but his data show that men actually hunt on only 31 percent of all possible days. 1 Additional data are available in Yellen's activity reports for 15 Bushmen camps (1977: appendix B), covering a 28-day period in May-June (cool-dry) and a 42-day period from December to March (warm-wet). No complete data are available for July to November. Yellen shows that men hunted on 47 percent of all days in May and June, and on 73 percent of all days in December through March (Table 13.1). In other words, the !Kung hunt more than five days a week on average in the hot, relatively wet summer, but only about two days a week in the cool, dry weather. The! Kung summer pattern is more like that of the Ache than it is like the !Kung winter pattern. This suggests that it may be water rather than heat stress that limits Kalahari hunters. 2 Only more quantitative information on seasonal hunting effort, water and temperature factors, and variations in prey distribution can clarify the seasonal costs and benefits of hunting. Empirical tests of hypotheses about specific trade-offs are clearly in order. Still, the very high hunting effort evident during the summer is sufficient to encourage the suggestion that !Kung men may well be working as hard and providing as much food as they safely can under the circumstances.


IV. Pre-State Foodways

TABLE 13.1. Hunting Effort Among !Kung Men in a Sample of 15 Camps

Dry Winter

May June

Camp number indentification
7.2 13 2 12 16

Adult man days of occupation
21 25 18 15 31

Dry Winter Totals
Wet Summer December January February


4.3 4.4 1.2 3.1 4.5 5 3.2 6 7.1

21 13 10 24 4 4
6 9

15 6

Wet Summer Totals
SOURCE: Tabulations from Yellen 0977: appendix B).


The Ache
Also low-latitude foragers, the Ache of Paraguay are not as well known as the !Kung. There are several references to the Ache (also called Guayaki) in historical accounts before the 1960s (see Metraux and Baldus 1963; O'Leary 1963), but the first modern ethnographic reports that were widely available are those of Clastres (1968, 1972), who studied two of the four living Ache groups. The data used here pertain to the northern Ache, who have come into unarmed contact with outsiders only during the past decade (Hill 1983). Full-time huntergatherers until that time, they currently live primarily at a Catholic mission


13. How Much Food Do Foragers Need?

Adult man days of hunting
13 9

Adult man days of gathering or honey collecting
1 7 5 1 7

% of

days men hunted

% of days men spent in food acquisition



4 4



8 4 15 4
2 6


9 15 5





(Chupa Pou) but continue to forage frequently in the nearby forest. These foraging trips have been described in reports that include figures on time costs and quantities of resources acquired, as well as analyses of the efficiency of Ache foraging strategies (Hawkes, Hill, and O'Connell 1982; Hill and Hawkes 1983), the seasonal pattern of food acquisition (Hill et al. 1984), the sharing of food resources (Kaplan et al. 1984; Kaplan and Hill 1985), and men's and women's time allocation (Hill et al. 1985; Hurtado et al. 1985). During the 10 years prior to their first peaceful contact with outsiders (1960-1970), the northern Ache probably numbered between 600 and 800 (Hill 1983). Their traditional range is an area of about 12,000 square kilometers


IV. Pre-State Foodways

between 54 and 56 degrees west and 24 and 25 degrees south. This is a wellwatered region with many rivers and streams. Most of the area is covered with tropical broad-leaf evergreen forest, which the Ache prefer to open grassland. Rainfall is quite unpredictable from month to month and year to year, although there is a statistically significant difference between wet and dry seasons. Annual average precipitation is approximately 1,600 millimeters. Fluctuations in temperature are much more regular, with an annual January maximum around 40°C and a July minimum of about -3°C. Ecology and climate are more fully described in Hill et al. (1984). The Ache take a wide variety of animal species. Among the most important are peccaries (Tajassu tajacu and Tajassu pecari), paces (Cuniculus paca), coatis (Nasua nasua), armadillos (Dasypus novemcintus), and capuchin monkeys (Cebus apella). They also exploit numerous plant products, especially those of the palm Arecastrum romanzoljianum, from which they take the fruit, the heart, and starch from the trunk. Fruits and honey are also major resources with insects providing a small but consistent component of the diet. Hunting in closed, sub-tropical forest, cross-cut by many small streams, the Ache are at much less risk of heat stress or dehydration than are the ! Kung. Here, too much water can limit hunting. Rain ruins bow strings, blurs tracks, makes the animals difficult to hear or see. The 15 percent of the study days on which men hunt less than two hours are days of heavy rain. Otherwise, and with surprisingly low variance (Hill et al. 1985), men hunt an average of seven hours a day, day after day, in every season. The simple proposition implicit in the discussion of the! Kung was that more food generally gives higher fitness benefits. Thus, acquisition will not stop when some "minimum needs" are met. Instead, work effort and procurement totals will be limited by the costs (and benefits) imposed by local ecological circumstances. Where these costs are lower, foraging investment should increase accordingly. The higher work effort of Ache men, throughout the seasons, is consistent with this. Two specific findings relevant here are that better hunters spend more time hunting (Hill and Hawkes 1983; Hawkes et al. 1985), and that better hunters have higher reproductive success (Kaplan and Hill 1985). Why then, if heat and water stress are unlikely and more food is generally better than less, should Ache women spend so little time gathering, an average of less than two hours per day (Hurtado et a1. 1985)? The answer to women's low food-related work effort may lie partly in the fact that Ache move camp nearly every day in pursuit of game. The women are responsible for carrying all household equipment and at least some of their children between camps. This suggests the hypothesis that women and their children do better by following the men than they would by establishing a more permanent camp in which gear and children could be deposited while they gathered food in the surrounding


13. How Much Food Do Foragers Need?

forest. Two implications of this are (1) that men should do better hunting on days when camp moves, and (2) that on days when camp is not moved, women's foraging effort should increase. In fact, if the days when camp is not moved are separated from the "normal activity" days, the difference in the time women spend acquiring and processing food is substantial. Their foraging time increases 77 percent on the rare days when camp does not move (93 versus 165 minutes). In addition to trading off foraging for carrying, Ache women face another problem. In an environment where the "safe area" for small children is never larger or more substantial than the temporary camps, the local exigencies of child care may constrain women's foraging. Women's collecting activities mav reduce the quality of their child care or significantly increase its cost where there is high probability of danger to children who are not attentively monitored. That child care does constrain foraging is indicated by the kind of foraging women do. They collect those items out of the larger set the Ache exploit whose acquisition is most compatible with child care. The trade-off is quantitatively demonstrated by the reductions in women's foraging efficiency (calories acquired per unit of time) when they are nursing, and also when the number of weaned dependents they have goes up (Hurtado et al. 1985). These data are consistent with the view that more foraging imposes a fitness cost on women, unmatched by its fitness benefit. What women "need" to collect is not set by nutritional minima, but by the fitness opportunity costs of foraging.

Even though provisional and illustrative, this argument about the determinants of variation in foraging effort and food acquisition among the Kung and Ache should be sufficient to suggest the potential utility of a sociobiological perspective in the analysis of human behavior. With respect to foraging, this perspective suggests that features of local ecology such as the presence of injurious flora and fauna, the distribution and availability of water and food resources, and seasonal changes in these will affect the costs and benefits associated with food acquisition. The expectation is that variable patterns are adaptations in the narrow sense-that is, they tend to maximize (inclusive) fitness. Whether arguments like this are ultimately compelling will depend on the results of further analyses of these and larger, more comprehensive data sets, especially those collected specifically for the purpose of testing sociobiologically derived hypotheses about behavioral variability. From a sociobiological perspective, patterns of behavior are expected to favor options that tend to have the highest fitness payoffs. Leisure, food, com-


IV. Pre-State Foodways

fort, security, health, power, affection, are not ends in themselves but means. More food generally means more grandchildren. If and when it does not, we expect acquisition to be limited accordingly. From this perspective many differences among modern foragers, and differences within and between local groups and across time among our Pleistocene ancestors, may be the result of fitness-enhancing adjustments to fairly immediate local constraints. It is not enough to see the rather different work effort and consumption patterns of the ! Kung and the Ache as the adventitious results of recent history. By taking advantage of the concept of adaptation in the strictest biological sense, we may be able to account for some of the variety in the present in the way that will help us understand the past. Among other things, it may prevent us from giving up on the simple answers too soon.

The foregoing description of the variation in foraging patterns between the !Kung and the Ache does not attend to the larger political and economic network in which both populations are embedded. This is an important general issue, independent of the merits of sociobiological perspectives. Can it be useful and clarifying to focus on the behavior of modem hunter-gatherers, and their local ecology, to the exclusion of wider historical contexts? In fact, given the special features of the modem world, can research on modern hunter-gatherers ever inform us in any general way about human foragers, especially in the prehistoric past? It is widely recognized that modem hunters are not pristine living relics of the Pleistocene. As Schrire (1980) has shown, the! Kung have alternately made their living as farmers and pastoralists as well as hunter-gatherers for at least the past four centuries and have often been the clients of more powerful neighbors, just as they are today. The Ache case is somewhat less clear but has some similar features. The northern Ache have been economically independent of its agricultural neighbors in the recent past, having only hostile contact with them until the establishment of local missionsponsored settlements in the early 1970s. At that time the Ache suffered enormous losses from disease. In 1978 the mission colony of Chupa Pou was created and is now home to more than two hundred Northern Ache. From here many continue to take extended foraging trips, sometimes longer than two weeks, into the surrounding forest, where they support themselves entirely by hunting and gathering in a manner they claim to be like their full-time foraging of a few years ago. The more distant past remains clouded. Their language is closely related to that of the surrounding horticultural Guarani Indians. It may be that they have ancestors in common, the Ache having been displaced by warfare to become refugee foragers sometime before the arrival of the Spanish in the 16th century. Such a pattern is probably not uncommon among lowland South American hunter-gatherers (Isaac 1977; Lathrap 1968).


13. How Much Food Do Foragers Need?

Some (notably Isaac 1977 and Schrier 1980) have argued that the historical circumstances of the !Kung and the Ache and other 20th-century foraging groups make them inappropriate sources of information and inference about "real" hunter-gatherers, especially those of the remote past. It is as though their recent history, with its episodes of disease, depopulation, and displacement, and its periods of dependence on horticulture or pastoralism, has somehow tainted them, made them "unrepresentative" as foragers. Moreover, it is argued that the social, political, economic, and ecological circumstances in which modern foragers are encountered are so much the product of colonialism and the reach of industrial capitalism, and so unlike the prehistoric past, that one cannot hope to learn anything useful from them apart from documenting the disastrous results of their contact with state societies. From that perspective studies of the foraging patterns of modem hunters that do not focus on the play of wider regional economic and political forces in their lives are, at best, deeply misguided. This conclusion can be challenged on several grounds. Most importantly, it implies little hope for a nomothetic understanding of human behavior. All societies differ from each other and change over time. Each occurs in a unique social and ecological setting. Any synchronic description slices out a unique set of events. It is the variability within and between societies and over time, the pattern in the differences observed ethnographically or inferred archaeologically, that we seek to explain. 'fhe descriptive convenience of typing societies-as "hunter-gatherers," for example-should not divert us into a search for archetypes in which the variability, the original object of inquiry, is cast aside as an imperfection, due to the damage of recent history, in some representations of the type. Any society anywhere has a recent history, and disease, depopulation, and displacement are not unique to the modem world. The question is, how do we go about explaining the variety? Although any particular instance of human behavior has some unique shape, we also expect that there are common processes that dictate this shape. How else do we account for the remarkable commonalities in the evolution of human behavior worldwide over the past hundred thousand years? If common processes are not at work in such phenomena as the independent near-simultaneous emergence of agriculture in several parts of the world at the beginning of the Holocene, or the subsequent, again more or less simultaneous and independent, emergence of state societies in a smaller number of cases, the coincidence can only be seen as astonishing. If we expect general principles to govern the larger pattern of temporal changes, we must also expect them to govern short-term patterns. T'he apparent uniqueness of modem circumstances must be accountable as well. We might use our sociobiological perspective to focus on the effects on modem foragers' behavior of particular features of their current state of dependence. The Ache mission shapes the opportunities available to individual Ache. We might seek to identify these alternatives, and the fitness-related costs and benefits of each, and make testable predictions about which of the array individual Ache will pursue. We might expect, for example, that variations in the amount of time individuals choose to spend at the mission or in the forest will depend on the respective fitness-related costs and benefits of each, and that these may vary with such factors as the age, sex, and health of those individuals, as well as with the changing mix of other individuals in either place, including potential mates, kin, and rivals. In this way


IV. Pre-State Foodways

a series of testable hypotheses can be developed about the effects of the mission on Ache behavior. The mission has changed the array of behavioral opportunities, but the process of choosing among them continues. In general, as in this paper, our focus of attention in studying the Ache has not been on the effects of the mission. Instead, we have focused on the behavior of the Ache observed away from the mission, in the forest. We have made the working assumption that certain behavioral strategies in the forest are shaped by the behavioral options in the forest. In other words, we assume that once Ache leave the mission to live entirely dependent on hunting and gathering for a week or two, some aspects of their behavior are shaped by the opportunity costs of foraging for different resources in different ways, and by the opportunity costs of foraging itself relative to other activities. We expect that people who spend significant portions of their time foraging in familiar habitats for their own subsistence in the company of kin, with mating opportunities largely limited to local groups, are likely to shape much of their behavior to these conditions. Such expectations are empirically vulnerable. If other factors-such as mission activities, features of the larger regional or world economic system, or historical circumstances not embodied in currently measurable options and constraints-overwhelmingly influence their choices, then correlations with local foraging conditions will be swamped. Surely many Ache (and ! Kung) behaviors depend entirely on the opportunities shaped by the larger state society in which they are embedded. But some behavior may be more powerfully governed by local ecology. If it is possible to account for much of the difference in work effort and food procurement by focusing on the opportunity costs set by such features as temperature and the distribution and abundance of water and wild plant and animal resources, we have a strong foundation for constructing expectations about foragers elsewhere, including behavior in particular times and places in the Pleistocene. As long as we construct and test theoretically motivated hypotheses about behavioral variety, we will check any tendency to give too much weight to the costs and benefits of immediate foraging options. It is important to know how much of the variation we can account for in this way. Unless we try, we may forfeit the richest line of evidence we have about human behavior in the Pleistocene.

james F. O'Connell is responsible for much of this paper and eludes co-authorship only at his firm wish. I thank Kim Hill, Nicholas Blurton jones, Kevin jones, Mark Cohen, Eric Charnov, Eric Smith, Eric Ross, Seymour Parker, Robert Anderson, and especially Hillard Kaplan for stimulating criticism and advice.

1. The alternative work opportunities for men at Dobe during Lee's 28-day sampling period may have made hunting significantly less attractive than during the same season in other years (Hill 1983).


13. How Much Food Do Foragers Need?

2. There are, however, some puzzling inconsistencies. For example, Lee (1979: 104-5), citing observations reported by Marshall (1976: 140), says that in the dry, hot spring (August-November), "the men say the heat makes the animals less wary and their movements are more predictable, so they are easier to stalk and kill. This fact is reflected in the kill statistics: October was a consistently good hunting month in terms of number of large animals killed." If water stress was a danger, one would have expected relatively low hunting effort during this period. Unfortunately, no data are available on the average number of days men actually hunt during this season.

References Cited
Blurton Jones, N., and R. Sibly 1978 Testing Adaptiveness of Culturally Determined Behavior: Do Bushmen Women Maximize Their Reproductive Success by Spacing Births Widely and Foraging Seldom? In Human Behaviour and Adaptation, N. Blurton Jones and V. Reynolds, eds., pp. 135-58. London: Taylor and Francis. Clastres, P. Ethnographie des Indians Guayaki (Paraguay-Bresil). Journal de La Societe des A mericanistes de Paris 57:9-6l. 1972 The Guayaki. In Hunters and Gatherers Today, M. B. Bicchieri, ed., pp. 138-174. New York: Holt, Rinehart and Winston. Cohen, M. 1977 The Food Crisis in Prehistory. New Haven: Yale University Press. Ember, C. 1978 Myths About Hunter-Gatherers. Ethnology 17:439-448. Harako, R. 1981 The Cultural Ecology of Hunting Behavior Among Mbuti Pygmies in the Ituri Forest, Zaire. In Omnivorous Primates: Gathering and Hunting in Human Evolution, R. S. O. Harding and G. Teleki, eds., pp. 499-556. New York: Columbia University Press. Harris, M. 1977 Cannibals and Kings: The Origins of Cultures. New York: Random House. 1983 Cultural Anthropology. New York: Harper and Row. Hawkes, K.; K. Hill; and]. F. O'Connell 1982 Why Hunters Gather: Optimal Foraging and the Ache of Eastern Paraguay. American Ethnologist 9: 379-98. Hawkes. K., and]. O'Connell 1981 Affluent Hunters? Some Comments in Light of the Alyawara Case. American Anthropologist 83: 622-26. Hawkes, K.; ]. F. O'Connell; K. Hill; and E. Charnov 1985 How Much Is Enough: Hunters and "Limited Needs." Ethology and Sociobiology 6:3-15.


IV. Pre-State Foodways

Adult Male Subsistence Strategies among Ache Hunter-Gatherers of Eastern Paraguay. Ph. D. dissertation, University of Utah. Hill, K., and K. Hawkes 1983 Neotropical Hunting Among the Ache of Eastern Paraguay. In Adaptive Responses of Native Amazonians, R. Hames and W. Vickers, eds., pp. 139-88. New York: Academic Press. Hill, K.; H. Kaplan; K. Hawkes; and A. M. Hurtado 1984 Seasonal Variance in the Diet of Ache Hunter-Gatherers in Eastern Paraguay. Human Ecology 12: 145-80. Hill, K., et a1. 1985 Men's Time Allocation to Activities Among Ache Hunter-Gatherers. Human Ecology 13:29-47. Howell, N. 1979 Demography of the Dobe Kung. New York: Academic Press. Hurtado, A. M.; K. Hawkes; K. Hill; and H. Kaplan 1985 Female Subsistence Strategies Among the Ache of Eastern Paraguay. Human Ecology 13:1-28. Isaac, B. The Siriono of Eastern Bolivia: A Reexamination. Human Ecology 1977 5: 137-54. Kaplan, H., and K. Hill 1985 Hunting Ability and Reproductive Success Among Male Ache Foragers: Preliminary Results. Current Anthropology 26: 131-33. Kaplan, H.; K. Hill; K. Hawkes; and A. M. Hurtado 1984 Ache Food Sharing. Current Anthropology 25: 113-15. Lathrap, D. The "Hunting" Economies of the Tropical Forest Zone of South Amer1968 ica: An Attempt at Historical Perspective. In Man the Hunter, R. Lee and 1. DeVore, eds., pp. 23-29. Chicago: Aldine. Leakey, R., and K. Lewin People of the Lake: Mankind and Its Beginning. Garden City, N. Y.: An1978 chor Press/Doubleday. Lee, R. B. What Hunters Do for a Living; or, How to Make Out on Scarce Re1968 sources. In Man the Hunter, R. B. Lee and 1. DeVore, eds., pp. 30-48. Chicago: Aidine. Kung Bushmen Subsistence: An Input-Out Analysis. In Environment 1969 and Cultural Behavior, A. P. Vayda, ed., pp. 47-79. Garden City, N.Y.: Natural History Press. Population Growth and the Beginning of Sedentary Life Among the 1972 !Kung Bushmen. In Population Growth: Anthropological Implications, B. Spooner, ed., pp. 329-42. Cambridge: MIT Press. The !Kung San: Men, Women and Work in a Foraging Society. Cam1979 bridge: Cambridge University Press.

Hill, K. 1983


13. How Much Food Do Foragers Need?

Lee, R. B., and 1. DeVore (eds.) 1968 Man the Hunter. Chicago: Aldine. Marshall, L. 1976 The !Kung ofNyae Nyae. Cambridge: Harvard University Press. Maynard Smith, ]. 1978 Optimization Theory in Evolution. Annual Review of Ecology and Systematics 9:31-56. Metraux, A., and H. Baldus 1963 The Guayaki. In Handbook of South American Indians, Vol. 1, J. Steward, ed., pp. 435-44. New York: Couper Square. O'Leary, T. 1963 Ethnographic Bibliography of South America. New Haven: Human Relations Area File Press. Pyke, G.; R. Pulliam; and E. L. Charnov 1977 Optimal Foraging: A Selective Review of Theory and Tests. Quarterly Review ofBiology 52: 137-54. Sahlins, M. 1968 Notes on the Original Affluent Society. In Man the Hunter, R. B. Lee and 1. DeVore, eds., pp. 85-89. Chicago: Aldine. 1972 Stone Age Economics. Chicago: Aldine. Schrire, C. 1980 An Inquiry Into the Evolutionary Status and Apparent Identity of San Hunter-Gatherers. Human Ecology 8:9-32. Smith, E. A. 1979 Human Adaptation and Energetic Efficiency. Human Ecology 7: 53-74. Tanaka, J. 1980 The San: Hunter Gatherers of the Kalahari-A Study in Ecological Anthropology. Translated by D. W. Hughes. Tokyo: University of Tokyo Press. Thomas, E. M. 1959 The Harmless People. New York: Knopf. Truswell, A. S., and J. D. L. Hansen 1979 Medical Research Among the Kung. In Kalahari Hunter-Gatherers, R. B. Lee and I. DeVore, eds., pp. 166-94. Cambridge: Harvard University Press. Williams, G. 1966 Adaptation and Natural Selection: A Critique of Some Current Evolutionary Thought. Princeton: Princeton University Press. Wilmsen, E. 1978 Seasonal Effects of Dietary Intake on Kalahari San. Federation ofAmerican Societies for Experimental Biology Proceedings 37: 65-72. Yellen, J. 1977 Archaeological Approaches to the Present: Models for Reconstructing the Past. New York: Academic Press.



Aboriginal Subsistence in a Tropical Rain Forest Environment: Food Procurement, Cannibalism, and Population Regulation in Northeastern Australia
In this chapter an indigenous, non-Westernized dietary pattern is analyzed as a contribution to the understanding of human food procurement and nutrition in tropical rain-forest environments. The example is drawn from northern Queensland, specifically the southeastern Cape York Peninsula, where an Australian Aboriginal population of foragers, fishers, and hunters continued to exist' unaffected by European influences, until the second half of the 19th century. The descriptive data on which the analysis is based are derived partly from 19th-century historical sources and partly from ecological and ethnographic fieldwork (Harris 1975, 1978). There is as yet little archaeological evidence relating to the antiquity and temporal development of the 19th-century dietary patterns (Campbell 1980; Harris 1978: 132; Hersfall 1983, 1984), but palynological investigation has demonstrated that tropical rain forest has been established in the area for approximately the last ten thousand years (Kershaw 1974, 1976).

Methodological Assumptions
The methodological assumptions on which this Australian case study rests accord with the adaptational and evolutionary perspective that underlies this volume as a whole. They can be enumerated as follows: 1. Persistent human dietary patterns consist of sets of behaviors involving food procurement, production, processing, and consumption that are them-


IV. Pre-State Foodways

selves part of larger subsistence systems composed of interacting ecological, physiological, demographic, technological, and sociopolitical variables. 2. Dietary patterns result from processes of evolutionary selection, the spatial and temporal dimensions of which need to be specified as precisely as possible. 3. Persistent (i. e., evolutionarily "successful") dietary patterns vary widely in relation to food availability and choice, but only within physiologically determined limits set by the basic nutritional requirements of humans for a dietary mix of proteins, carbohydrates, fats, and essential minerals and vitamins sufficient to provide energy and to sustain growth and reproduction. 4. A dietary pattern is definable in terms of behavioral norms in food-related activities that are common to, and the result of ecophysiologically optimizing behavior by, a group as a whole, but . . . 5. Group dietary behavior may incorporate socially prescribed food preferences and avoidances that take effect within or between groups and are keyed to differences in age, sex, status, or ethnicity. In analyzing the example presented in this Chapter, I focus on how the foodgetting and food-using activities of the Aboriginal population meshed with ecological, demographic, and social variables that form part of the subsistence system as a whole, giving special attention to how basic nutritional requirements were satisfied. Less attention is paid to the temporal development of the dietary patterns described because there is as yet so little chronological evidence available; and ethnograhic information on socially prescribed food preferences and avoidance is also too meager to allow more than causal reference to them.

The Cultural Context
Prior to the penetration of northeastern Queensland by European prospectors, loggers, and settlers in the second half of the 19th century, the mountains, plateaus, and coastal lowlands between Cardwell and Cairns were clothed in structurally complex and floristically diverse rain forest. The pre-European distribution of rain forest coincided closely with the east coast area, which receives more rainfall annually (1, SOO-ca. 4,000 millimeters) than any other part of Australia, although even here the winter months Gune-November) are relatively dry. In the mid-19th century, the rain-forest region of approximately 11,660 square kilometers was occupied, according to Tindale (1974:164-90), by 12 Aboriginal tribal groups whose territories were, with one exception, restricted to coastal and inland rain-forest habitats. There is uncertainty over


14. Aboriginal Subsistence in a Tropical Rain Forest Envirorunent

the precise number and distribution of "rain-forest tribes" (Dixon 1976; Peterson 1976), but there is little doubt that the Aboriginal populations of the region were in many ways physically and culturally distinct from the tribal groups of the open-woodland savanna habitats farther inland. Nineteenth-century European observers were impressed by the small stature and slender limbs of the rain-forest Aborigines, and Tindale and Birdsell (1941), who worked with a remnant population on the Atherton Plateau in 1938, identified the 12 tribal groups of the rain-forest region and argued that they constituted a distinctively pygmoid racial type. This led to a controversy about their racial affinities and putative role in the initial peopling of Australia, which lies outside the scope of this paper, but later linguistic studies by Dixon (1972, 1976) have tended to support the view that the rain-forest tribes were clearly differentiated from their neighbors, although at least five distinct languages were spoken in the region and linguistic links existed with peripheral groups. It is in their material culture, however, that the rain-forest tribes exhibit their greatest distinctiveness. Several items of their material culture are unique within Aboriginal Australia and directly reflect features of the rain-forest habitat. These include dome-shaped, rainproof, thatched huts occupied mainly in the wet season (December-May), but lasting from year to year; bark cloth hammered from the inner bark of fig trees (Ficus pleuroearpa), which was used to make blankets as well as containers both for carrying water and honey and for leaching bitter yams; baskets made from lawyer cane (Calamus australis) and rush (Xerotes longlfolia), which were used as sieve bags for leaching toxic nuts; huge wooden swords and shields, the latter made from the flange buttresses of fig trees and painted with intricate designs when a boy was initiated into manhood (McConnel 1935); and, of potential archaeological significance, very large stone axes used, for example, to lop off the fig tree buttresses, and two unique types of specialized nut-processing stone tools. The first of these was a large anvil stone pitted with small, spherical depressions, accompanied by a small hammer stone, used to crack open the exceptionally hard nuts of the Queensland almond (E laeoearpus banero/tiz). The other was an ovate or rectangular grooved slab (the morah), associated with a smaller crushing stone (the moogz), used to macerate nut kernels, especially those of the yellow and black walnuts (Beilsehmiedia banero/tii and Endiandra palmerstoniz). In addition to these items of material culture, several non-material aspects of rain-forest culture also appear to have been distinctive-for example, the holding of fighting corroborees to settle disputes, the practice of partial mummification prior to cremation, and, of particular interest in the context of this volume, cannibalism. The last trait was frequently-and avidly-commented upon by early European observers, but it is difficult to determine from the historical sources what role it may have played in pre-European rain-forest society. Its possible significance is discussed below.


IV. Pre-State Foodways

Food Procurement
It is possible to reconstruct in outline the pre-European pattern of food procurement of those rain-forest tribes who occupied the inland basaltic, granitic, and metamorphic plateaus and mountains, as opposed to the alluvial coastal lowlands, and who did not apparently have direct access to the coast and its resources. Four main historical sources provide information on the Aborigines of this area, all dating from the 1880s. They are the records left by Christie Palmerston (1883, 1886, 1887), a pioneer prospector; Carl Lumholtz (1889), a Norwegian zoologist and ethnographer; Archibald Meston (1889, 1904), who led two scientific expeditions to the Bellenden Ker Range; and Walter Roth (1901-10), the pioneer ethnographer who became the Queensland Government's first Northern Protector of Aborigines in 1897. Useful additional information on Aboriginal subsistence can be gleaned from a series of newspaper articles written by Robert Johnstone (1903-5), who served in the Native Police and traveled extensively in the rain-forest region in the 1870s and 1880s. On the basis of these historical sources, together with limited information from living descendants of the rain-forest tribes and ecological observations in the few remaining tracts of uncleared forest, it is possible to construct a broad picture of pre-European subsistence. The most detailed descriptions of Aboriginal food procurement are contained in Meston's reports on his two expeditions to the Bellenden Ker Range. They are worth quoting in full because they provide the most comprehensive picture of subsistence in the rain forest:
Their food is chiefly vegetarian, varied occasionally by the flesh of the wallaby, the tree-climbing kangaroo, fish, birds, eggs, and three or four varieties of opossums. The koa nut and other large nuts not yet botanically named, are the chief articles of diet. Some of the nuts and roots they eat are poisonous in their raw state, and these are pounded up and placed in dilly bags in running water for a couple of days to have the poisonous principle washed out. Of edible nuts of various kinds they have an unlimited supply. In pursuit of tree-climbing animals they display an agility probably unsurpassed in the world, and probably not rivalled by any other Australian black. It would severely tax the reader's credulity to describe how these natives take a vine and run up the tallest trees, walk on to others across the branches, and descend sometimes a conside